Respondents Brief — Microsoft Corp. v. AT & T CORP.

No. 05-1056

IN THE

Suprene Court of the United States

MICROSOFT CORPORATION,

Petitioner,

AT&T CORP.,

Respondent,

ON WRIT OF CERTIORARI TO THE

UNITED STATES COURT OF APPEALS

FOR THE FEDERAL CIRCUIT

BRIEF FOR RESPONDENT

LAWRENCE. LAFARO SETH P. WAXMAN

THOMAS A. RESTAINO Counsel of Record

AT&T CORP. WILLIAM G. MCELWAIN

One AT&T Way JONATHAN FE. NUECHTERLEIN

Scdminster, NJ 07921 MARK C. FLEMING

(QOS) S8B-1TS50 WILMER CUTLER PICKERING

HALE AND DORR LLP

L875 Pennsylvania Ave., N.W.

Washington, DC 20006

(202) 663-6000

QUESTIONS PRESENTED

1. Whether digital software code—-an intangible se-

uence of “1’s” and “0’s”—may be considered a “component{]

of a patented invention” within the meaning of 35 U.S.C.

§ 271(f); and, if so,

2. Whether Microsoft’s transmission of such code

abroad constitutes the “suppl[y]” of such a component within

the meaning of that provision.

(i)

Te

shah ®

PY grant a4 mit he His ail, Se oa as a

eater yee, Me nee

7 y “ : f

“ask ey

a

met ee ~*~

ale

Al Ae, * |

Pe fs a ' va

~~ : P} ’ ~ % at ae ee

ln ' ‘3 ‘ . _<

fag® .

La

a ee

ki

apes

TABLE OF CONTENTS

Page

QUESTIONS PRESENTED...............sccscssssssssessssscecseseressecseenes i

TABLE OF AUTHORITIEBG...............cccsccossscessssescercecsossssesenes Vv

PRELIMINARY STATEMENT ............csccsssssessssssesssssesssseeses 1

PE A I cnrcninvisorvissnbniensicinntdnnesinnesiniiimecsesdiiniaianmniigitetcinitanents 4

A. The Physical And Non-Physical Layers Of

ET 4

i ee pesitirtcestaccsscpticsintiinteiaintstnncciinniiti 7

C. Microsoft’s Acknowledged Infringement

Sr re ee crrtrttntttintincicinnanstammmennnninn 8

D. The District Court Judgment. .........cccssecsssesseeees 10

E. The Court Of Appeals Judgment.............ssssseeeees 1]

SUMMARY OF ARGUMENT 00. eececcssessecseeesenreneneseees 13

IT iviscniiccicntitcinnssniatemnctininitannnnpinnitiatemnneniolitiaiin 15

I. INTANGIBLE OBJECT CODE IS A “COMPO-

NENT” OF A PATENTED INVENTION. ......ccsscsssssessseeseseees 16

A. If Microsoft Is Correct That The First

Question Presented In The Petition Is Not

Properly Before The Court, Certiorari

Should Be Dismissed As Improvidently

B. Section 271(f) Encompasses All “Compo-

nents” Of An Invention, Including Intan-

gible Components Such As Object Code ..............+ 19

1. The term “component” encompasses

both physical and non-physical ase

Ce iia crccenntitttncnceenicitinesiinicstintnbitnsinaninagnin 19

(iii)

iv

TABLE OF CONTENTS—Continued

2. Object code is an essential component

of software technology even though it

must be combined with physical com-

ponents tO FUNCION.............ccceceeserseeseeenerees

II. MICROSOFT “SUPPLIED” INTANGIBLE OBJECT

CODE FROM THE UNITED STATES FOR COM-

BINATION WITH OTHER COMPONENTS

ABROAD

A. Microsoft’s Arguments On The “Supply”

B.

Issue Presuppose The Validity Of Its Er-

roneous Argument On The “Component”

Microsoft’s Invocations Of Legislative His-

tory Are Unavailing.............ccscssssscessssssssssseseees

III. MICROSOFT IDENTIFIES NO POLICY BASIS

FOR DISREGARDING THE PLAIN MEANING OF

THE STATUTORY TEXT

A. Far From Vindicating The “Technology-

Neutral” Purposes Of Section 271(f), Mi-

crosoft’s Position Would Repeal That Pro-

vision For Software Components. ..............000+

The Presumption Against Extraterritori-

ality Is Inapplicablle ...............scsscsssesssscesesenesncens

Microsoft’s “Outsourcing” Arguments Are

EE IE ESET CSE a Cw OTE OR

CONCLUSION

Vv

TABLE OF AUTHORITIES

CASES

Page(s)

Addamax Corp. v. Open Software Foundation,

Inc., 152 F.3d 48 (1st Cit, 1998)........csscccsrcsrcsesersesecsersnees 21

American Trim, L.L.C. v. Oracle Corp., 383 F.3d

SED Gas Cie. BB DG) ncentccapnscccinsternescschninianinsinenctetesciainisisueti 29

Apple Computer, Inc. v. Formula International,

Inc., 594 F. Supp. 617 (C.D. Cal. 1984)......cccccccssssseeseenes 40

Bayer AG v. Housey Pharmaceuticals, Inc.,

340 F.3d 1367 (Fed. Cir. 2008) ...........cccescesesesseseeesenees 23, 49

Brown v. Duchesne, 60 U.S. (19 How.) 183 (1857) ............00 43

Computer Associates International, Inc. v. Altai,

Inc., 982 F.2d 693 (2d Cir. 1992)..........-cescessscsecsesersesesses 37

Deepsouth Packing Co. v. Laitram Corp., 406 U.S.

BO COE ccncncccecsncscceninneinseuidunsesibainiansianenpianianttaath 14, 35, 42

Deere & Co. v. International Harvester Co.,

710 F.2d 1551 (Fed. Cir. 1983)...........cscssssssssssesrsssssesersees 34

Diamond v. Diehr, 450 U.S. 175 (1981)......csecsessessesessseereseees 23

Dowagiac Manufacturing Co. v. Minnesota Moline

Plow Co., 235 U.S. 641 (1915).......csccssccssesssesssesessesesseneees 43

Dresser-Rand Co. v. Virtual Automation, Inc., "

361 F.3d 831 (Sth Cir. 2004) .........ccsssesesessessesessesssseesseseses 29

EEOC v. Arabian American Oil Co., 499 U.S, 244

CBE cenicsersccssetasissntenmcimissenitavisiabubssiiitaiadtiianiaaibabiiinass 42, 44

Eolas Technologies Inc. v. Microsoft Corp.,

399 F.3d 1325 (Fed. Cir.), cert. denied, 126 S.

Ce, A ID sieitsttiesnsecectinenieivlsaeiiiisidipiaii eidiitinaitad passim

F. Hoffmann-La Roche Ltd. v. Empagran S.A.,

GOB TE. BD GO cecenccecnscenericininantictzeeninanintatninditie 44, 45

Foley Bros., Inc. v. Filardo, 336 U.S. 281 (1949) ......cscsesees 42

General Motors Corp. v. Devex Corp., 461 U.S. 648

CRIED erosccecersissssovintsininpesssicineneivecnasiteiniatinpiinaligiasinteisiainian 34

Globetrotter Software, Inc. v. Elan Computer

Group, Inc., 362 F.3d 1367 (Fed. Cir, 2004)... 21

Hartford Fire Insurance Co. v. California, 509 U.S.

Bee Cie cccectscnncpnicnagupitebenvianpiinicsinneitieesniinimdisbiiiindiinpaiida 45

vi

TABLE OF AUTHORITIES—Continued

Page(s)

ISC-Bunker Ramo Corp. v. Altech, Inc., 765 F.

i ENS E> I ccnccenicnicissnsentinestevstcenenemnnnsnsenanenten 6

Microsoft Corp. v. Commissioner, 311 F.3d 1178

STII iindtiiticiinbinetnatataninenintantsenncinemmnepemmaniien 9, 32

Oncale v. Sundowner Offshore Services, Inc.,

SE Uecker dich ieeptateiiteenetnerineaieatmemenenen 35

Pellegrini v. Analog Devices, Inc., 375 F.3d 1113

(Fed. Cir.), cert. denied, 543 U.S. 1003 (2004)................ 27

Response of Carolina, Inc. v. Leasco Response,

Inc., 5387 F.2d 1307 (Sth Cir, 1976) .........ccccccececeeeeeesesees 4,21

Rogers v. United States, 522 U.S. 252 (1998) ..........cccccceseseeee 17

Sheldon v. Metro-Goldwyn Pictures Corp., 106 F.2d

45 (2d Cir. 1939), aff'd, 309 U.S. 390 (1940) ..........ccceeee 47

Smith v. United States, 507 U.S. 197 (1998) .........cccccsceeseeeees 43

Specht v. Netscape Communications. Corp.,

ee Ee Ge Ee See ccctencscnsnssnnteseminnssnsnntecnenseneespienses 29

Steele v. Bulova Watch Co., 344 U.S. 280 (1952) .............0000 45

United States v. Microsoft Corp., 253 F.3d 34 (D.C.

8 EE ee ae 21, 29

United States v. Spelar, 338 U.S. 217 (1949).......ccccccccseseeeees 43

Waymark Corp. v. Porta Systems Corp., 245 F.3d

Se, CII Dcemnemntnenrecsintntnneansenaresmnsemmamen 36, 41

Williams v. Taylor, 529 U.S. 420 (2000) ...........cccccccsecsesseeeeeees 12

Wireless Agents, L.L.C. v. Sony Ericsson Mobile

Communications AB, No. 3:05-CV-0289-D,

2006 WL 2239112 (N.D. Tex. Aug. 4, 2006) ..............0000 21

STATUTES AND RULES

EE EN ee 30

a since nbieeensenencennnennenetonemnmtetens 42

26 U.S.C. § 927(a)(2)(B) (repealed 2000) epnteensintasiengeeneiatiantis 32

Kk Ee ee 43

EE 22, 26

Sk ee 22

8 EE passim

ae Cie Se Oren eneescanessszsnscnsepnmnencscsrscinanvensnsnensenseennnsnntennanenaesnee 34

vii

TABLE OF AUTHORITIES—Continued

Page(s)

Sa Ee cnnintesnencnntinennenentpiittessttepeniettagriiuenmeneiniiadiaiis 22

35 U.S.C. § 292...........cc00000 sindelpaninitingiselatientanatuaiasaniad 22

BP Te ele eseccictrectinsiananetnniniidvetntititinasiniteenmeneinmnnenee 17

LEGISLATIVE MATERIALS

Amendment in the Nature of a Substitute to H.R.

2795, the “Patent Act of 2005”: Hearing Before

the Subcomm. on Courts, the Internet, and In-

tellectual Property of the House Comm. on the

Judiciary, 109th Cong., Ist Sess. (2005).............ccceseeees 49

Committee Print Regarding Patent Quality Im-

provement: Hearing before the Subcomm. on

Courts, the Internet, and Intellectual Property

of the House Comm. on the Judiciary, 109th

CUR, BIR IER, Ge eenenenssninmenninns 48

130 Cong. Rec. 28,069 (1984), reprinted in 1984

aes CIID tactesenennnconnnsnnincemmetinnpnnatennteneemenies 49

Subcommittee on Courts, the Internet, and Intel-

lectual Property of the House Committee on

the Judiciary, Committee Print: Patent Act of

SERS, F BD Cia. 06, Ba ccccnsccnssesnssestoennssecseecsciemesemasens 30

Patent Reform Act of 2006, S. 3818, 109th Cong.

Fy Gren scennsnpnemenmnnmmeeenemnpmniemecin 37

INTERNATIONAL MATERIALS

Agreement on Trade-Related Aspects of Intellec-

tual Property Rights, Including Trade in Coun-

terfeit Goods, Apr. 15, 1994, in Marrakesh

Agreement Establishing the World Trade Or-

ganization, Annex 1C, 33 I.L.M. 1197 (1994)..............0 45

OTHER AUTHORITIES

Benkler, Y., From Consumers to Users: Shifting

the Deeper Structures of Regulation, 52 Fed.

Communications L.J. 561 (2000) .............cccccsceesseceeeeesenees 25

TABLE OF AUTHORITIES—Continued

Page(s)

Chisum, D., Normative and Empirical Territorial-

ity in Intellectual Property: Lessons from Pat-

ent Law, 37 Va. J. Int’l] L. 603 (1997). 2 47

Langlois, R., Modularity in Technology and Or-

ganization, 49 J. Econ. Behav. & Org. 19 (2002) ............. 6

Lessig, L., The Future of Ideas: The Fate of the

Commons in a Connected World (2002)... —

Mairs, J., VPNs: A Beginner’s Guide (2002)... 26

Microsoft Computer Dictionary (5th ed. 2002)............... 19, 20

Oppliger, R., Security Technologies for the World

Se Cte Ot eh cirenctresesenticnnenmnaieicces 26

Peterson, J. & A. Silberschatz, Operating System

I cttintitdintearenansreantniaitinenienitasmeemenenen 20

Plato, Cratylus, quoted in G.S. Kirk & J.E. Raven,

The Presocratic Philosophers (1971) om

Pressman, R., Software Engineering: A Practitio-

ner’s Approach (6th ed. 2005)............... 5, 20, 22, 24, 25, 37

Pressman, R., Software Engineering: A Practitio-

a errnttterictiintetiieimnnieneian 23

Silberschatz, A., P. Galvin & G. Gagne, Operating

System Concepts (7th ed. 20085) ......... passim

Stern & Gressman, Supreme Court Practice (8th

ed. 2002) ee 17

Szyperski, C., Component Software: Beyond Object-

Oriented Programming (2d ed. 2002) ...........cececceceeeseceeee 5

Tyson, Jeff, How Computer Memory Works, at

http://computer.howstuffworks.com/computer-

memory.htm (visited Jan. 23, 2007).............. ... 40

Tyson, Jeff & Dave Coustan, How RAM Works, at

http://computer.howstuffworks.com/ram.htm

CO SI, Fite a eencmnscnsepressesesemeemmmnemeeemesmesen 40

United States Patent & Trademark Office, Manual

of Patent Examining Procedure § 2106 wide ed.

2003)... sevens —_ — 11

ix

TABLE OF AUTHORITIES—Continued

Page(s)

United States Patent & Trademark Office, Manual

of Patent Examining Procedure (8th ed., 5th

rev., 2006)

a 21

ES ee a

PE eceaitenrectuvennnanesintensinncteninapiteininning 21

Walters, E. Garrison, The Essential Guide to Com-

I Gore crensccssnsesinsnssenniemeccmtmmansemmpianinmnayee 24, 40

Webster’s Third New International Dictionary

(1981) —— 19, 24, 28

‘es

fre SG

P G5 ta

2 ot ra

: ott Pn

a

ol er

Page

IN THE

Supreme Court of the United States

No. 05-1056

MICROSOFT CORPORATION,

Petitioner,

v.

AT&T CORP.,

Respondent.

ON WRIT OF CERTIORARI TO THE

UNITED STATES COURT OF APPEALS

FOR THE FEDERAL CIRCUIT

BRIEF FOR RESPONDENT

PRELIMINARY STATEMENT

Section 271(f) of the Patent Act makes it an act of pat-

ent infringement to “suppl[y] ... from the United States”

either—

all or a substantial portion of the components of a

patented invention . .. in such manner as to actively

induce the combination of such components outside

the United States in a manner that would infringe

the patent if such combination occurred within the

United States,

35 U.S.C. § 271(f)(1) (emphasis added), or

any component of a patented invention that is espe-

cially made or especially adapted for use in the in-

vention and not a staple article or commodity of

commerce suitable for substantial noninfringing

use, ... knowing that such component is so made or

adapted and intending that such component will be

2

combined outside the United States in a manner

that would infringe the patent if such combination

occurred within the United States,

id. § 271(f)(2) (emphasis added). At issue in this case is

whether these provisions grant any meaningful protection

for inventions practiced through computer software.

During all relevant periods, AT&T held the patent to a

sophisticated speech coding and decoding technology. Al-

though AT&T licensed this technology to many companies

on reasonable terms, Microsoft wished to exploit AT&T’s

innovation for free. Microsoft now concedes that it infringed

AT&T’s patent for years by supplying domestic computer

manufacturers with software object code that enabled per-

sonal computers in the United States to make use of AT&T’s

invention. Microsoft also supplied the same object code to

foreign manufacturers for the same purpose, kept close

track of how often they loaded the code onto new computers,

and charged them a licensing fee each time they did so. Mi-

crosoft nonetheless contends that it may escape liability un-

der Section 271(f) because the code was carried on different

physical media to its ultimate destination in foreign com-

puters, and because Microsoft did not supply the medium

that carried the code on the last leg of its journey. The dis-

trict court and court of appeals rejected this argument, held

Microsoft liable, and explained that Microsoft’s position

would strip Section 271(f) of any significance in the software

context.

Microsoft sought certiorari on two questions. First, it

asked the Court to decide “{wJhether digital software code—

an intangible sequence of ‘1’s’ and ‘0’s’"—may be considered a

‘component({] of a patented invention” within the meaning of

Section 271(f). Pet. i (questions presented). Second, Micro-

soft asked the Court to decide whether, if the answer to that

first question is yes, its transmission of the object code

abroad constituted the “supply” of such a component. Mi-

crosoft rightly gave top billing to the first of these two ques-

tions. Whether “an intangible sequence of ‘1’s’ and ‘0’s”” (id.)

can be a “component” of an invention for Section 271(f) pur-

3

poses is the central issue in this case and logically precedes

the “supply” issue.

Microsoft now flips the order of its discussion and ad-

dresses the “supply” issue first, but that turns the logical

structure of this case on its head. Microsoft’s arguments on

that issue presuppose that the Court has already accepted

Microsoft’s dubious claim, deferred to the end of its brief,

that intangible object code cannot be a “component” of an

invention. When Microsoft finally does address the meaning

of “component,” it argues for the first time that the very

question on which it sought certiorari—whether “an intan-

gible sequence of ‘1’s’ and ‘0’s’ ... may be considered a com-

ponent” (Pet. i.)—is not properly presented after all, because

of how Microsoft reads various stipulations in the district

court. This is a perplexing argument, because if it had merit

and were preserved, certiorari would need to be dismissed

as improvidently granted.

If the Court does not dismiss the writ, the answer to the

first question presented is straightforward. Microsoft’s ob-

ject code is not just a “component,” but the key component of

the foreign-made devices in terms of their ability to practice

AT&T’s invention. Although that object code must be com-

bined with physical components to form such a device, it is

plainly a component of that device, just as a unique collection

of intangible words is a component of any book bearing the

title Moby-Dick, even though those words, too, must be

combined with ink and paper before the book can be read.

Microsoft would read into Section 271(f) an “implied re-

quirement that a ‘component’ be physical in nature” (Pet. Br.

42 n.14 (emphasis added)), but that position lacks any basis

in the statutory text and improperly conflates the physical

and non-physical layers of computer technology. And if that

position were adopted, it would read Section 271(f) out of the

Patent Act for virtually any invention practiced by the use

of software, a result Congress did not intend. Finally, Mi-

crosoft plainly “supplied” this object code from the United

States to foreign computer manufacturers, with the intent

that those companies would pay Microsoft a royalty each

4

time they combined that code with other components to

form devices that would infringe AT&T’s patent if made or

used in the United States.

Microsoft and its allies—including the amicus trade as-

sociations that they fund—have long sought to persuade

Congress to repeal Section 271(f) altogether. Having so far

failed in these legislative efforts, the same alliance urges this

Court to strip Section 271(f) of all meaning in the software

industry. But the law is the law as it stands today. Under

any faithful reading of that law, Microsoft has infringed

AT&T’s patent and is liable under Section 271(f).

STATEMENT

A. The Physical And Non-Physical Layers Of Computer

Technology

At a basic level, computer systems are made up of

hardware and software “components.” A. Silberschatz, P.

Galvin & G. Gagne, Operating System Concepts 3 (7th ed.

2005) (“Silberschatz”) (stating that the “components” of a

computer system include “hardware,” “operating system,”

and “applications programs”). These components are com-

bined to achieve the “fundamental goal of computer sys-

tems,” which is to “execute user programs and to make solv-

ing user problems easier.” /d. at 5.

The physical parts of a computer are known as “hard-

ware.” Today’s hardware consists of the physical materials

inside the “box”—such as the central processing unit

(“CPU”), memory cards, hard drives, and circuitry—and ex-

ternal physical parts such as the video monitor, keyboard,

mouse, printer, cabling, and removable devices for storing

information, such as compact disks (“CDs”), floppy disks,

and magnetic tapes. See, e.g., Response of Carolina, Inc. v.

Leasco Response, Inc., 537 F.2d 1307, 1326 (5th Cir. 1976).

Computer hardware is not useful unless “programmed”

to perform a particular function. A program directs a com-

puter’s CPU to open or close numerous electrical switches at

particular times, thereby creating electrical signals that

cause the various hardware elements to perform desired

5

tasks. See, e.g., R. White, How Computers Work 53 (8th ed.

2006).

In modern computing technology, the hardware is pro-

grammed by software in the form of object code. Object

code—also called “machine language”—is expressed as a

precise sequence of binary digits (1s and Os) that turn par-

ticular switches within a computer’s microchip circuitry “on”

and “off.” White, swpra, at 87.' The same object code can be

stored or transported in any of a number of different physi-

cal containers. For example, the same sequence of Is and 0s

can be represented in the arrangements of indentations

(“pits”) and unindented spaces (“lands”) on the surface of a

plastic CD; in rapid bursts of light within a fiber-optic cable;

or in the orientations of magnetic fields on a computer’s hard

drive. But regardless of where a program is stored, it can be

executed by a CPU only once the Is and Os have been trans-

ferred from their storage location to a computer’s random-

access memory, or “RAM,” where they are represented as

patterns of electrical charges on a RAM chip. See id. at 49.

Today’s software technology is highly “modular,” which

means that software engineers can develop and market their

products for use on many different types of computer hard-

ware and in conjunction with many other types of software.

Software is thus routinely referred to as a “component” of a

larger computer system,’ and software engineers work on

such components without needing to worry about the physi-

' Human programmers normally write software not as object code,

but in one of several programming languages such as C++, BASIC, or

FORTRAN. Software in that format is known as “source code” and must

first be translated into object code through a process called “compiling” in

order to function successfully on a computer. See, e.g., White, supra, at 94.

? See, e.g., Silberschatz, supra, at 3; see also, e.g., R. Pressman, Soft-

ware Engineering: A Practitioner’s Approach 125 (6th ed. 2005) (“In the

software context, a component could be a computer program, a reusable

program component, a module, a class or object, or even a programming

language statement.”); C. Szyperski, Component Software: Beyond Ob-

ject-Oriented Programming 10 (2d ed. 2002) (“The distinguishing proper-

ties of software are of a mathematical rather than a physical nature.”).

6

cal details of how the code they write will be expressed on

any particular machine. Such modularity allows the soft-

ware industry to achieve enormous efficiency and flexibility

by “breaking up a complex system into discrete pieces—

which can then communicate with one another only through

standardized interfaces within a standardized architecture—

[to] eliminate what would otherwise be an unmanageable

spaghetti tangle of systemic interconnections.” R. Langlois,

Modularity in Technology and Organization, 49 J. Econ.

Behav. & Org. 19, 19 (2002).

Computer scientists thus describe a modern computer

system as consisting of several different “layers” of modular

technology. See Silberschatz, supra, at 72. The bottommost

layer consists of a computer system’s physical hardware,

such as the CPU and RAM, as well as the other physical de-

vices used to store and transmit data. The higher layers

consist of software code. “Each layer is implemented with

only those operations provided by lower-level layers. A

layer does not need to know how these operations are im-

plemented; it needs to know only what these operations do.

Hence, each layer hides the existence of certain data struc-

tures, operations, and hardware from higher-level layers.”

Id. at 60.

The layered nature of software and hardware is im-

mensely important to the structure of the computer market-

* See, e.g., ISC-Bunker Ramo Corp. v. Altech, Inc., 765 F. Supp.

1310, 1318 (N.D. Ill. 1990) (“Software is commonly developed to be used in

layers.”). Microsoft’s Windows software is itself divided into various lay-

ers. For example, Microsoft’s Windows XP system is built on a “hard-

ware-abstraction layer” or HAL, which “manipulates hardware directly,

isolating the rest of Windows XP from hardware differences among the

platforms on which it runs.” Silberschatz, supra, at 787. This lower layer

in the operating system enables Windows to be “moved from one hard-

ware architecture to another with relatively few changes” (a feature

known as “portability”). /d. Windows also includes a large number of

higher-layer “applications” programs, such as a calculator, clock, Internet

browser, and—particularly relevant here—voice-manipulation programs

such as NetMeeting and Sound Recorder, which enable the practice of

AT&T’s patented technology. J.A. 16-17.

7

place. Software developers such as Microsoft need not also

build computers on which to run their programs, and com-

puter firms like IBM or Dell need not develop software tai-

lored to their specific systems. Instead, as long as certain

compatibility standards are met, software object code can be

developed and run on different manufacturers’ computer

systems. While both object code and hardware are needed

for a computer to function properly, they may be—and fre-

quently are—developed, advertised, and purchased sepa-

rately. See Silberschatz, supra, at 836 (“(U]sers can choose

and upgrade hardware to match their budgets and perform-

ance requirements without needing to alter the applications

they run.”). Accordingly, software developers and consum-

ers understand that a particular piece of software—such as

Windows, WordPerfect, or TurboTax—refers to a particular

program, regardless of whether the program’s object code is

burned onto on a compact disk, saved on the magnetic plat-

ters of a computer’s hard drive, transmitted over a fiber-

optic cable, or executed on a RAM chip.

B. AT&T's Invention

In 1981, two scientists at Bell Laboratories (then part of

AT&T), Dr. Bishnu Atal and Mr. Joel Remde, filed a success-

ful application for a patent on a pioneering advance in digital

speech compression. Modern telecommunications systems

generally transmit speech by converting it into digital data

or “code”; the code is transmitted to its destination, where it

is then “decoded” back into a speech signal. Dr. Atal and

Mr. Remde invented a novel technique that greatly en-

~ hanced the quality of the speech signal heard at the destina-

tion while decreasing the amount of data that needed to be

transmitted. This invention—disclosed and claimed in U.S.

Reissue Patent 32,580 (“the 580 patent”), which AT&T held

until it expired in 2001—is widely recognized as a landmark

in telecommunications technology and has won many pres-

tigious awards. Ct. App. J.A. 509-510. |

As relevant here, the ’580 patent claims an apparatus

comprising means for generating coded speech signals from

audible voice sounds and for receiving those coded signals

8

and converting them back into audible voice sounds. Supp.

J.A. 18-19 (’580 patent, cls. 29, 40-41). From the outset, Dr.

Atal and Mr. Remde recognized that the invention would be

practiced by writing software that, when installed on a com-

puter, would enable the computer to code and decode speech

signals in the disclosed manner. Indeed, Dr. Atal wrote such

software himself and appended excerpts of the source code

to the ’580 patent. Jd. at 12-16.

In 1996, the International Telecommunications Union,

an organization responsible for the promulgation of interna-

tional standards for the telecommunications industry, rec-

ommended this speech compression technique as an industry

standard. Many companies—including cellular telephone

manufacturers, software developers, videoconferencing pro-

viders, and Internet companies whose products involved

speech transmission—licensed the technology from AT&T to

ensure that their products were compatible with the recom-

mended standard. As Microsoft notes (Br. 3), AT&T’s tech-

nology, discovered over 25 years ago, is still widely used to-

day “in mobile phones and personal computers to achieve

high-quality reproduction of digitally recorded speech.”

C. Microsoft’s Acknowledged Infringement Of AT&T's

Patent

Microsoft, a U.S.-based software developer, has long

used AT&T’s speech-compression technique by including in

its Windows operating system certain object code that, once

installed on a compatible computer and loaded into RAM,

enables the computer to perform the coding and decoding

(“codec”) functions claimed in the ’580 patent. Microsoft de-

velops, tests, and debugs its Windows software, including

speech codecs, in the United States. Pet. App. 22a. Micro-

soft then markets the finished software product in the

United States and abroad, where it is pre-installed on com-

* AT&T also asserted that Microsoft infringed two other claims cov-

ering methods for coding and decoding speech. Jd. at 16, 19 (580 Patent

cls. 2, 42).

9

puters manufactured by other companies and also sold as

separately packaged software. Microsoft collects a license

fee for every computer sold abroad that contains its Win-

dows software. See Microsoft Corp. v. Commissioner, 311

F. 3d 1178, 1181 (9th Cir. 2002).

Microsoft refused to license AT&T’s patented technol-

ogy despite repeated requests (J.A. 17), and in 2001 AT&T

sued Microsoft for patent infringement. AT&T asserted

several bases of infringement, two of which Microsoft con-

ceded after AT&T presented its case-in-chief to the jury.

First, Microsoft stipulated that it directly infringed AT&T’s

patent under Section 271(a)° by making infringing devices in

the United States: namely, by installing the Windows object

code on its own computers during the process of developing,

testing, and debugging the software. Pet. App. 42a; Pet. Br.

4n.1. Second, Microsoft stipulated to liability under Section

271(b)® for actively inducing other parties to infringe

AT&T’s patent—specifically, by sending the Windows object

“code to U.S. computer manufacturers (also called original

equipment manufacturers or “OEMs”) with the specific in-

tent that they install Windows on their computers and

thereby make devices that infringe AT&T’s patent. Pet.

App. 42a; Pet. Br. 4.

At issue here is Microsoft’s liability for its provision of

the Windows object code to computer manufacturers outside

the United States. Microsoft provides that object code to

foreign manufacturers from the United States in the same

manner as it provides the code to U.S.-based manufacturers:

by sending it on a CD known as a “golden master disk” or in

an encrypted electronic transmission directly to the manu-

facturer. Pet. App. 23a, 45a 74 4-5, 7 (stipulation). The code

is then transferred onto the hard drives of foreign-

*“(W]hoever without authority makes . . . any patented invention,

within the United States ... , infringes the patent.” 35 U.S.C. § 271(a).

° “Whoever actively induces infringement of a patent shall be liable

as an infringer.” 35 U.S.C. § 271(b).

10

manufactured computers, which are sold in foreign countries

as computers containing Microsoft’s Windows operating sys-

tem. Id. at 45a.’ Microsoft stipulated that it specifically in-

tends that the object code it sends will be installed into those

computers. /d. at 46a ¥ 9. It is undisputed that, if per-

formed in the United States, the assembly and use of the

computers containing that object code would infringe

AT&T’s patent.

AT&T claimed that, through this scheme, Microsoft

supplied a “component” of a patented invention from the

United States in a manner that infringed AT&T’s patent un-

der 35 U.S.C. § 271(f)(1) and (2). Microsoft moved for sum-

mary judgment, arguing that software cannot be a “compo-

nent” under the statute because it is “intangible.” Pet. App.

24a. In its reply brief in the district court, Microsoft raised

the additional argument that the Windows object code in-

corporated into the foreign-manufactured computers had not

been “supplied from” the United States because the code

was replicated abroad before installation. Jd. at 24a-25a.

D. The District Court Judgment

The United States District Court for the Southern Dis-

trict of New York rejected Microsoft’s argument that the

Windows object code cannot be a “component” under Section

271(f). The court recognized that “[t]he object code or soft-

ware that is contained on each golden master disk or trans-

mitted electronically, as opposed to the golden master disk

or method of encrypted transmission itself, is at the heart of

the parties’ dispute.” Pet. App. 29a. The court also noted

that, despite Microsoft’s argument that software code is “in-

tangible,” it is nonetheless a component of a patentable ap-

paratus when used “in conjunction with a physical structure

such as a computer memory.” /d. at 30a-3la (quoting

’ Microsoft also delivered its object code to foreign companies called

“replicators,” which transferred the Windows object code exactly as it

was supplied from the United States to other storage media for sale or

delivery to customers. Pet. App. 23a, 45a { 6.

11

United States Patent & Trademark Office, Manual of Pat-

ent Examining Procedure (“MPEP”) § 2106, at 2100-13 (8th

ed. 2003)). Moreover, the court added, neither the statutory

text nor the legislative history supports limiting the term

“component” to physical machines or structures, nor did

Congress ever suggest that it meant to exclude software

components from Section 271(f). Id. at 3la. The court also

noted that object code is actually “incorporated into the end-

product” and therefore rejected Microsoft’s claim that object

code should be treated as design information, assembly in-

structions, or data generated from a patented process. /d. at

34a.

The district court further held that Microsoft “supplied”

the Windows object code from the United States. Pet. App.

35a. Microsoft’s contrary argument, the court explained,

ignored the undisputed fact that “the object code is origi-

nally manufactured in the United States.” Jd. The court

also deemed it significant that Microsoft itself had “acknowl-

edged that if individual disks with the infringing Windows

object code were sent abroad for incorporation into each for-

eign-assembled computer (rather than one golden master

disk), Microsoft would be liable for infringement under Sec-

tion 271(f).” Jd. at 36a n.7. The Court found that there was

no basis under “the letter and intent of the statute” to dis-

tinguish between that situation and this case. /d. at 35a-36a.

Based on these conclusions, Microsoft stipulated to a

judgment of infringement and entered into a settlement

agreement with AT&T. Pet. App. 42a-43a. That agreement

preserves Microsoft’s right to appeal the district court’s de-

cision regarding Section 271(f), and prescribes different dol-

lar amounts that Microsoft must pay AT&T depending on

the outcome of the appeal.

E. The Court Of Appeals Judgment

The court of appeals affirmed. At the outset, it noted

that the first question presented—“whether software may

be a ‘component’ of a patented invention under § 271(f)"—

had been answered in Eolas Technologies Inc. v. Microsoft

12

Corp., 399 F.3d 1325 (Fed. Cir.), cert. denied, 126 S. Ct. 568

(2005), in which the court had held that Section 271(f) is not

limited to “patented ‘machines’ or patented ‘physical struc-

tures.” Pet. App. 4a (quoting 399 F.3d at 1339).

With respect to the second question—whether the Win-

dows object code in the foreign-made computers had been

“supplied” from the United States—the court of appeals

sought to discern the “ordinary, contemporary, common

meaning” of the word “supply” in the context of software

distribution. Pet. App. 6a (quoting Williams v. Taylor, 529

U.S. 420, 431 (2000)). Like the district court, the court of

appeals rejected Microsoft’s argument that Section 271(f)

liability could attach where “each disk . . . is shipped and in-

corporated into a foreign-assembled computer,” yet not

where a single disk was sent with the intent that the object

code it contained be replicated and incorporated into each

computer, thereby “saving material, shipping, and storage

costs.” Jd. at 7a. The court of appeals also rejected Micro-

soft’s analogy to design instructions, since the Windows ob-

ject code was shipped “ready for installation on a computer

to form an infringing apparatus” and did not constitute “in-

structions to foreign software engineers for designing and

coding Windows.” /d. at 8a.

Judge Rader dissented. Although he agreed that soft-

ware was a “component” under Section 271(f) (Pet. App. 11a)

and recognized that the remaining question was whether

“intangible software components” were supplied from the

United States (id. at 13a), he argued that liability should not

attach because “the master disk” was not itself incorporated

into the foreign-made computers (id. at 16a). Judge Rader

did not explain the apparent discrepancy between his initial

finding that intangible software could be a “component” and

his later assumption that the “component” in this case was a

physical master disk. Judge Rader also believed that impos-

ing liability on Microsoft in this case improperly gave “ex-

traterritorial effect to U.S. patent laws” (id. at 17a), even

though Microsoft was held liable solely for actions it took

within the United States.

13

SUMMARY OF ARGUMENT

1. This case turns on the answer to the first question

presented in Microsoft’s petition for certiorari: “[w]hether

digital software code—an intangible sequence of ‘l’s’ and

‘0’s’—may be considered a ‘component{] of a patented inven-

tion’ within the meaning of Section 271(f)(1).” Pet. i (ques-

tions presented). Having persuaded this Court to review

that question of general application, Microsoft now advances

an interpretation of the record that, in its view, precludes

the Court from resolving the question. It is unclear what

Microsoft hopes to achieve by this tactic. If its argument on

this point were both preserved and meritorious (it is nei-

ther), the proper course would be not to rule for Microsoft,

but to dismiss the writ of certiorari as improvidently

granted. ;

If the Court does not dismiss the writ, the first question

presented in the petition should be answered in the affirma-

tive. The term “component” refers to any part of a larger

system, whether tangible or intangible, and object code is

plainly a component of any computer system that practices

AT&T’s invention. Indeed, computer scientists routinely

speak of software “components” that are independent of any

particular physical-layer medium that may be used to store

or transmit those components at any given moment. Of

course, object code must be combined with physical-layer

components to create a patentable machine, but that does

not make it any less a component in its own right. Micro-

soft’s contrary argument both ignores the plain meaning of

the term “component” and collapses the distinction, central

to the computer industry, between higher (code) and lower

(hardware) layers of computer technology.

2. An affirmative answer to the first question presented

compels an affirmative answer to the second question:

whether Microsoft “supplie[d]” object code for combination

with physical-layer components abroad to produce devices

that, if assembled in the United States, would directly in-

fringe AT&T’s patent. That is exactly what Microsoft did.

Microsoft’s contrary view assumes the correctness of its ar-

14

gument on the first question presented (addressed second in

Microsoft’s brief) that the relevant “component” is the

physical medium containing the object code, not the code

itself, which remains the same regardless of how it is physi-

cally embodied at any given moment. Microsoft’s assump-

tion is false for the reasons discussed.

Nor is there merit to Microsoft’s argument that, be-

cause Congress’s enactment of Section 271(f) was prompted

by this Court’s decision in Deepsouth Packing Co. v.

Laitram Corp., 406 U.S. 518 (1972), the provision should ap-

ply only in cases where, as in Deepsouth itself, a U.S. firm

ships physical components for assembly abroad. If Congress

had meant to confine Section 271(f) to physical components,

it would have said so in the text. And because that provision

indisputably applies in contexts far removed from the facts

of Deepsouth, it would be nonsensical to rely on those facts

as a basis for reading artificial limitations into the statutory

language.

3. Microsoft’s appeal to principles of “technological neu-

trality” is highly ironic, because those principles cut strongly

against Microsoft’s position. If, as Microsoft argues, only

physical items containing object code can qualify as “compo-

nents,” Section 271(f) would have vanishingly narrow appli-

cation to the software industry. For example, as Microsoft

now appears to acknowledge, its position would insulate it

from liability even if it directly sent an individual CD con-

‘taining the relevant object code for each foreign-made com-

puter, because end users normally transfer such code from

the pits and lands of the CD to magnetic properties on each

computer’s hard drive. Microsoft cannot seriously deny that

its position would amount to a near-total repeal of Section

271(f) for the software industry—an outcome Microsoft has

doggedly (but so far unsuccessfully). pursued in Congress.

Under current law, however, Microsoft’s position is a most

implausible reading of this statute, as even Microsoft con-

cedes that Congress had no intent to treat the software in-

dustry differently from any other industry.

15

Microsoft’s separate invocation of the presumption

against extraterritoriality is untenable in several respects.

First, the presumption is simply inapposite here, because

Section 271(f) renders Microsoft liable not for “extraterrito-

rial” conduct, but for domestic conduct: shipping software

object code from the United States with the requisite intent.

Second, the presumption cannot supersede statutory text

that speaks directly to a provision’s geographic scope, as

Section 271(f) does. Third, it would be particularly inappro-

priate to apply a judicial presumption to narrow the scope of

a statute that Congress enacted for the sole purpose of over-

coming the judicial application of that very presumption in a

prior case. Fourth, it would make no policy sense to apply

that presumption here, as there is no basis for concern about

any actual conflict with the prerogatives of any foreign sov-

ereign.

Finally, Microsoft’s other policy arguments quarrel not

with the application of Section 271(f) in this context, but

with Congress’s decision to enact that provision in the first

place. Congress considered and rejected the arguments that

foreign patent protections are sufficient to protect U.S. in-

ventors and that enactment of Section 271(f) would create

undue incentives for certain types of businesses to move

their operations offshore. Congress, no*. this Court, should

perform any fine-tuning of that policy judgment.

ARGUMENT

This case presents two questions: first, whether intan-

gible object code can constitute a “component” of a patented

invention for purposes of Section 271(f); and second, whether

Microsoft “supplied” such a component from the United

States to its foreign business partners. These two questions

should be asked and answered in that logical order. It is im-

possible to determine whether Microsoft “supplied” the

relevant “component” from the United States without first

deciding what the relevant “component” is—which, as the

Solicitor General observes, is the “logically antecedent”

question of the two (U.S. Cert. Br. 19).

16

Perhaps recognizing that the case turns on identifying

the “component” at issue—and that its position on that ques-

tion is weak—Microsoft inverts the questions presented and

assumes throughout Part I of its brief that the Court has

already agreed with Microsoft that a Section 271(f) “compo-

nent” must be a particular physical thing embodying object

code, rather than the object code itself. Because that ap-

proach begs the primary question in this case, we address

the questions in the order in which they are presented in Mi-

crosoft’s petition.

I. INTANGIBLE OBJECT CODE Is A “COMPONENT” OF A PAT-

ENTED INVENTION

A. If Microsoft Is Correct That The First Question Pre-

sented In The Petition Is Not Properly Before The

Court, Certiorari Should Be Dismissed As Improvi-

dently Granted

In its petition for certiorari, Microsoft invited this Court

to review an issue of general application: “[wJhether digital

software code—an intangible sequence of ‘1’s’ and ‘0’s’—

may be considered a ‘component{] of a patented invention’

within the meaning of Section 271(f)(1).” Pet. i (emphasis

added). That is the precise issue addressed by both lower

courts, which likewise conceptualized the disputed “compo-

nent” as the intangible “object code contained on the golden

master disks,” not as the golden master disks themselves.

Pet. App. 24a (district court); see also id. 4a (court of appeals

holding that Section 271(f) is not limited to “physical struc-

tures” (quoting Eolas Technologies Inc. v. Microsoft Corp.,

399 F.3d 1325, 1339 (Fed. Cir.), cert. denied, 126 S. Ct. 568

(2005))." As the district court confirmed, “the heart of the

parties’ dispute” has always involved “(t]he object code or

software that is contained on each golden master disk or

transmitted electronically, as opposed to the golden master

disk or method of encrypted transmission itself.” Jd. at 29a.

* See also Pet. App. 13a (Rader, J., dissenting) (noting that this case

involves “intangible software components”).

17

Having obtained certiorari, Microsoft now argues that

this case does not present the first question after all. It con-

tends that the parties stipulated that “the ‘components’ in

issue were the golden master disks and the encrypted

transmissions,” not the intangible object code that they

transmitted, and that the lower courts should have decided

the case on that basis. Pet. Br. 10, 34-35. This is a bizarre

gambit. If this argument were correct and preserved, “the

record [would] not fairly present” the question of general

application on which Microsoft sought certiorari, Rogers v.

United States, 522 U.S. 252, 259 (1998), and the appropriate

course would be to dismiss the writ as improvidently

granted. See id.; Stern & Gressman, Supreme Court Prac-

tice 329 (8th ed. 2002) (citing cases). This Court did not

grant certiorari to correct the lower courts’ supposed mis-

understanding of a case-specific stipulation of facts.

If, however, the Court elects to decide this case on the

merits, Microsoft’s argument about that stipulation is nei-

ther preserved nor correct. First, Microsoft waived this

new argument by failing to present it either to the court of

appeals or in its petition for certiorari. Before the court of

appeals, Microsoft challenged the district court’s conclusion

that “the intangible Windows object code, as distinct from a

golden master disk or other software media upon which the

. software information may be stored, was a component of the

patented computer systems supplied from the United

States.” Microsoft Ct. App. Br. 8. And Microsoft never gave

this Court any inkling at the certiorari stage that it would

later try to sabotage the Court’s consideration of what Mi-

crosoft called a “recurring question of vital importance.”

Pet. 11.

This Court’s rules admonish even a respondent, in op-

posing certiorari, that “[a]Jny objection to consideration of a

question presented based on what occurred in the proceed-

ings below, if the objection does not go to jurisdiction, may

be deemed waived unless called to the Court’s attention in

the brief in opposition.” S. Ct. Rule 15.2. A fortiori, a peti-

tioner may not pull the rug out from under this Court’s con-

18

sideration of an issue after persuading this Court that the

issue has broad general significance and that the case pre-

sents a sound vehicle for resolving it.’

Second, Microsoft’s interpretation of the stipulation is

without merit. The stipulation states: “AT&T alleges, and

Microsoft disputes, that the computer systems assembled

abroad with the foreign replicated object code contain ‘com-

ponents’ that were supplied by Microsoft from the United

States.” Pet. App. 46a (emphasis added). The statement on

which Microsoft fastens—that Microsoft did not supply a

component “other than the ‘golden master disks’ and the en-

crypted transmissions of Windows object code,” id. at 47a

(emphasis added)—merely focused the analysis on the “Win-

dows object code” as transmitted via the golden master

disks and electronic transmissions. It does not state, as Mi-

crosoft now contends (Br. 34), that the “[pJhysical [mJedia”

are the only “components” at issue in this case. The district

court confirmed AT&T’s understanding of the stipulation in

its final judgment (to which Microsoft also stipulated), hold-

ing that “the golden master disks and the encrypted trans-

missions of Windows object code contain ‘components’ sup-

plied from the United States by Microsoft” and that “the

computer systems assembled abroad with the foreign-

Against this backdrop, it is the height of irony for Microsoft to ac-

cuse AT&T of engaging in a “duck-and-dodge tactic” by discussing the

question on which certiorari has now been granted (Pet. Br. 34). Re-

markably, Microsoft chides AT&T for including in its final certiorari-stage

brief a description of the first question presented that is taken directly

from Microsoft's own petition and the Solicitor General's brief in support

of certiorari. Compare Pet. Br. 33 (criticizing AT&T for characterizing

the first question presented as whether “intangible 1s and 0s,’ a binary

sequence of numbers that ‘lacks physical existence,” can qualify as a

“component” (quoting AT&T Second Supp. Br. 1, 4)) with Pet. i (charac-

terizing question presented as whether “an intangible sequence of ‘1’s’ and

‘O's’ ... may be considered a ‘component™) and U.S. Cert. Br. 8-9

(addressing Microsoft’s argument that “software cannot be a ‘component’

of a patented invention because it is ‘intangible information” and because

“the concept of the Windows software lacks physical existence” (quoting

Pet. 15-17)).

19

replicated Windows object code that was installed from the

golden master disks or the ence:~’pted transmissions contain

‘components’ that were supplied by Microsoft from the

United States.” Pet. App. 42a | 4 (emphasis added).

B. Section 271(f) Encompasses All “Components” Of

An Invention, Including Intangible Components

Such As Object Code

1. The term “component” encompasses both

physical and non-physical parts of a system

The Windows object code consists of many types of pro-

grams, some of which enable a general-purpose computer to

practice AT&T’s invention. Microsoft has stipulated that it

directly infringed AT&T’s patent under Section 271(a) by

making and using infringing computer devices in the United

States while developing, testing, and debugging the Win-

dows code on Microsoft-owned computers. See Pet. App.

42a; Pet. Br. 4 n.1. And Microsoft also stipulated that it

unlawfully induced the infringement of AT&T’s patent un-

der Section 271(b) by sending the Windows object code to

U.S. computer manufacturers for inclusion in computers sold

in the United States, in essentially the same way that it sent

the same code to foreign manufacturers for inclusion on com-

puters abroad. See Pet. App. 42a; Pet. Br. 4.

The primary question in this case is whether that object

code is a “component” of the foreign-made computers that

practice AT&T’s patented technology, such that Microsoft’s

transmissions of the object code to foreign manufacturers

infringed AT&T’s patent under Section 271(f). Under any

plain-language interpretation of that term, the answer is

yes.

“Component” means “a constituent part” or “ingredi-

ent.” Webster’s Third New International Dictionary 466

(1981) (“Webster’s Third Int'l’). The term broadly encom-

passes not just the physical, but also the non-physical, parts

of a composite system or device. See id. (citing, as example

of usage, “the essential [components] of Kantian philoso-

phy”); see also J.A. 34 (Microsoft Computer Dictionary 116

20

(5th ed. 2002)) (defining “component” as a “discrete part of a

larger system or structure”).

Accordingly, the word “component” is routinely used to

describe software independent of any physical format,

whether as part of a computer system consisting of both

hardware and software or as part of a larger non-physical

software program.” Microsoft itself uses the term that way

in its own publications and patents.'' The U.S. Patent and

Trademark Office’s Manual of Patent Examining Procedure

See, ¢.g., Silberschatz, supra, at 3 (including among the “compo-

nents” of a computer system “the hardware, the operating system, [and]

the application programs” (emphasis in original)); J. Peterson & A. Sil-

berschatz, Operating System Concepts 1 (1983) (same use of “compo-

nents”); Pressman, supra, at 125 (“In the software context, a component

could be a computer program, a reusable program component, a module, a

class or object, or even a programming language statement.”). Indeed, an

entire academic literature has arisen to explore how software engineers

design “component software” for use and reuse within larger software

systems—a discipline that refutes Microsoft's effort to limit the word

“component” to “physical” material (Br. 42 n.i4). See, e.g., Pressman, su-

pra, at 7 (“A software component should be designed and implemented so

that it can be reused in many different programs.”); id. at 815-816 (defin-

ing the provess of “component-based software engineering” as “the design

and construction of computer-based systems using reusable software

‘components”’).

'' See, e.g., J.A. 34 (Microsoft Computer Dictionary 116 (5th ed.

2002)) (defining “component” as a “[ajn individual modular software rou-

tine that has been compiled and dynamically linked, and is ready to use

with other components or programs”); J.A. 29 (Microsoft Windows prein-

stallation guide) (describing as “Additional components” any “hardware,

drivers, applications, and so on that you want to preinstall” (emphasis

added)). One of Microsoft’s patents reproduces a fragment of source code

that it describes as having “three main components,” the functions of

which are described without reference to any particular physical medium.

Supp. J.A. 62-63 (U.S. Patent No. 6,738,773, col. 6 1. 58 to col. 7, 1. 41).

Other Micrvsoft patents likewise refer to software components regardless

of any particular physical embodiment. See, e.g., id. at 29 (U.S. Patent No.

6,725,262, col. 5, ll. 23-28) (referring to “executable software components”

that “reside at various times in different storage components of the com-

puter”); id. at 50 (U.S. Patent No. 6,727,917, col. 3, ll. 11-14) (stating that

“input/output components” include “software”).

21

employs the same usage,” as do numerous judicial deci-

sions.’ Several of Microsoft’s own amici thus concede that

Microsoft is wrong on the first question presented and that

intangible software indeed qualifies as a “component” for

purposes of Section 271(f). See, e.g., American Intellectual

Property Law Ass’n Br. 4-11; Houston Intellectual Property

Law Ass’n Br. 6-8; Intellectual Property Professors Br. 1;

Professor Edward Lee Br. 3 n.4."*

'2 See, e.g., MPEP § 2106.01 at 2100-17 (8th ed., 5th rev., 2006) (dis-

cussing “data structures and computer programs which impart functional-

ity when employed as a computer component”); id. § 2106.01, at 2100-18

(referring to “computer software and hardware components”); id.

§ 2161.01, at 2100-163 (referring to patent applications that claim elements

that are “partially comprised of a computer software component”); id.

§ 2164.06(c), at 2100-198 (referring to computer systems “which include a

computer as well as other system hardware and/or software compo-

nents”).

'? See, e.g., Globetrotter Software, Inc. v. Elan Computer Group,

Inc., 362 F.3d 1367, 1370 (Fed. Cir. 2004) (stating that license agreements

may “prevent the software purchaser from using different components of

a suite of programs on different computers simultaneously”); United

States v. Microsoft Corp., 253 F.3d 34, 93 (D.C. Cir. 2001) (“A justification

for bundling a component of software may not be one for bundling the

entire software package ... .”); Addamaz Corp. v. Open Software Found.,

Inc., 152 F.3d 48, 49 (ist Cir. 1998) (“[Sjecurity software is a component

that can be used with the operating system to restrict outside access to

sensitive information.”); Response of Carolina, Inc. v. Leasco Response,

Inc., 537 F.3d 1307, 1326 (5th Cir. 1976) (stating that a computer system

can be “broken into three components: the computer hardware, the oper-

ating system (systems software) and the applications programs (applica-

tion software)”); Wireless Agents, L.L.C. v. Sony Ericsson Mobile Com-

muns. AB, No. 3:05-CV-0289-D, 2006 WL 2239112, at *2 (N.D. Tex. Aug.

4, 2006) (discussing patent infringement “by devices that use non-physical

components, e.g. software”’).

‘* The Solicitor General (Br. 10) supports the court of appeals’ con-

- ¢lusion that “software can be a component of a patented invention,” but

then claims that the “component” at issue is the “physical copy of the

software installed on a particular computer.” But the court of appeals

expressly rejected Microsoft’s argument that components must be “physi-

eal,” as the Solicitor General himself acknowledges. See Pet. App. 4a

(stating that “components” are not limited to “physical structures” (quot-

ing Eolas, 399 F.3d at 1339)); U.S. Br. 6 (quoting the same language). The

22

Microsoft nonetheless continues to advocate what it

candidly describes as an “implied requirement that a ‘com-

ponent’ be physical in nature.” Pet. Br. 42 n.14 (emphasis

added). But that “implied” limitation would contradict the

plain text of the statute. If Congress had so intended, it

could have confined Section 271(f) to the supply of “tangible”

components of an invention, as Microsoft prefers (Pet. i), or

“physical” components, as the Solicitor General prefers (Br.

13), or “elements” of a patent claim, as amicus Eli Lilly pre-

fers (Br. 12). Congress certainly knew how to limit patent-

law provisions in those respects, having framed other sec-

tions of the Patent Act by reference to claim elements or

physical matter.’ But Congress chose not to impose such

limitations on the scope of Section 271(f), and for good rea-

son. Intangible “code ... is not only a component,” but “the

key part” of virtually any invention practiced through soft-

ware. Eolas, 399 F.3d at 1339 (emphasis added). Without

the object code, the foreign-made computers would be col-

lections of useless hardware; they certainly would not be

able to perform the complex task of coding and decoding

speech signals. See, e.g., Pressman, supra, at 2 (software

“delivers the computing potential embodied by computer

hardware”).

Indeed, as discussed in Point III.A below, if object code

as such did not qualify as a “component” of such products,

Section 271(f) would have no meaningful application to the

software industry. Congress did not intend that result.

Software was central to many patentable inventions when

Solicitor General nowhere explains why the relevant “component” should

be limited to a “physical copy” when the court of appeals—which the So-

licitor General contends answered the first question presented “correctly”

(U.S. Br. 7)—held to the contrary.

'> See, e.g., 35 U.S.C. § 101 (providing for patenting of a “machine,

manufacture, or composition of matter”); id. § 112 4 6 (referring to an

“element” of a patent claim, as well as “structure” and “material”); id.

§ 287(a) (providing for marking of “any patented article” by “fixing

thereon the word ‘patent””); id. § 292(a) (prohibiting “mark[ing] upon, or

affixing] to, ... any unpatented article” the word “patent”).

23

Congress enacted Section 271(f) in 1984."° And as Microsoft

itself agrees, “there is no indication that Congress meant to

treat software any differently” from other technology for

purposes of Section 271(f). Pet. Br. 8-9 (internal quotation

marks and ellipsis omitted)."’

2. Object code is an essential component of soft-

ware technology even though it must be com-

bined with physical components to function

Microsoft contends (Br. 38) that object code itself, apart

from any physical manifestation, cannot be a “component”

because it is not “readable and executable by a computer” in

that state. That is a non sequitur. Of course object code

must be combined with a physical component—such as a CD,

hard drive, or memory chip—before it can be transmitted or

used. But in that sense it is no different from any other

component addressed by Section 271(f), which may have no

utility on its own but yields a novel and useful device when

'® See, e.g., Diamond v. Diehr, 450 U.S. 175, 187 (1981) (holding that

a device using a computer program may be patentable); Jn re Comstock,

481 F.2d 905, 909-910 (C.C.P.A. 1973) (same); R. Pressman, Software En-

gineering: A Practitioner’s Guide 1 (2d ed. 1987) (noting that Business

Week article entitled Software: The New Driving Force, published on Feb-

ruary 23, 1984, appeared “about ten years too late,” given that “software

has surpassed hardware as the key element to the success of many busi-

nesses, products, and systems” and “is often the key factor that differen-

tiates”). Indeed, the original patent application for AT&T’s technology

was filed in 1981, and the patent issued in 1984.

'’ Microsoft’s reliance on Section 27 l(g) for its contrary interpreta-

tion of Section 271(f) (Br. 42 n.14) is meritless. Section 271(g) bases in-

fringement liability not on the shipment of a component, but on the impor-

tation of “a product which is made” by a U.S.-patented process. 35 U.S.C.

§ 271(g) (emphasis added). Although the words “product” and “made”

have been held to reach physical products only, see Bayer AG v. Housey

Pharms., Inc., 340 F.3d 1367, 1377 (Fed. Cir. 2003), neither word appears

in Section 271(f). And while Section 271(g) envisions that a “product . . .

made” outside the United States might become a “trivial and nonessential

component of another product,” the fact that one physical “product” might

become a “component” of another physical “product” under Section 271(g)

does not remotely suggest that intangible object code cannot be a “com-

ponent” of a “patented invention” for purposes of Section 271(f).

24

combined with other components. Congress nowhere sug-

gested that Section 271(f) applies only to “components” that

are independently useful or novel before their combination

with other components. Cf. U.S. Br. 12 (“The non-

patentability of software code standing alone has no bearing

... on whether software can be a component of a patented

invention under Section 271(f).” (emphasis omitted)).

Microsoft similarly argues (Br. 11) that object code itself

cannot be a “component” for Section 271(f) purposes on the

theory that, if it lacks physical existence, it is not “capable of

being ‘combined’ with other components” and, indeed, “can-

not be combined with anything.” See also id. at 42-44; ef.

U.S. Br. 15. This is nonsense. In ordinary language, “com-

bine” means “to cause (as two or more things or ideas) to

mix together,” as exemplified by such locutions as “combin-

ing the language of the gutter with ideas of undoubted

worth” and “his talents and looks [combined] got him the

job.” Webster’s Third Int'l 452 (definitions of “combine” and

“combined”). It is perfectly natural to speak of combining

intangible object code with physical components such as a

hard drive or CD to make software technology work within a

variety of computer systems. See, ¢.g., Pressman, supra, at

124 (stating that the various elements of a computer-based

system, including software and hardware, “combine in a va-

riety of ways”); Autodesk Br. 7 (stating that “intangible”

software code is “combined outside the United States”).

Microsoft’s contrary argument ignores not only the

plain meaning of the text, but the basic structure of modern

computer technology. As discussed in the Statement, supra,

today’s computer systems consist of conceptually distinct

layers of technology, some tangible and some intangible. In

a computer system, “hardware is the bottom layer, and

software sits on top.” E. Garrison Walters, The Essential

Guide to Computing 135 (2001); see also Silberschatz, supra,

at 60. Similarly, the public Internet consists not just of

“physical layer” equipment in the form of wires and routers,

but also (among other things) “a ‘logical’ or ‘code’ layer—the

code that makes the hardware run,” including the intangible

25

“protocols that define the Internet.” L. Lessig, The Future

of Ideas: The Fate of the Commons in a Connected World 23

(2002) (citing Y. Benkler, From Consumers to Users: Shift-

ing the Deeper Structures of Regulation, 52 Fed. Communi-

cations L.J. 561, 562-563 (2000)). The Internet exists be-

cause engineers in various disciplines have succeeded in

combining the technologies on these distinct layers into a

single communications system.

The personal computer industry operates on the same

basic principle. Computer manufacturers such as Dell or

HP, and CD manufacturers such as Philips, make physical-

layer devices for storing or transporting object code. These

companies can modify or upgrade their technology’s ability

to handle any given sequence of 1s and 0s without having to

worry about precisely which sequences of 1s and 0s their

products will handle or for purposes of what application.

Similarly, software developers such as Corel or Adobe may

write computer programs without fretting the details of pre-

cisely how the Is and 0s of the programs’ object code will be

expressed in the pits and lands of a given CD, in the mag-

netic storage devices of a given personal computer, or in the

bursts of light in a given telecommunications carrier’s fiber-

optic transmissions. Such independence of the physical layer

is possible because “[sloftware is a logical rather than a

physical system element.” Pressman, supra, at 5 (emphasis

added)."*

Here, the use of AT&T’s invention requires a combina-

tion of several different components. On the physical layer,

it requires hardware, including a RAM chip and a CPU. On

a non-physical level, it requires the use of intangible object

code, which (once loaded into RAM) tells a computer’s CPU

'® See also Silberschatz, supra, at 60 (“Each layer is implemented

with only those operations provided by lower-level layers. A layer does

not need to know how these operations are implemented; it needs to know

only what these operations do. Hence, each layer hides the existence of

certain data structures, operations, and hardware from higher-level lay-

ers.”).

26

how to manipulate particular speech signals. Of course, the

code component alone is not itself a “process, machine,

manufacture, or composition of matter” and must therefore

be combined with the physical-layer components before an

inventor can obtain a patent. 35 U.S.C. § 101." Nonetheless,

code remains not only a component of that invention, but by

far the most important component from an intellectual prop-

erty perspective. Only when combined with object code can

the foreign-made computers (often a collection of mostly

commodity parts) become an invention that is “new and use-

ful.” Id.”

By analogy, the unique series of words that constitutes

Moby-Dick, while intangible, retains its independent iden-

tity as such no matter how it is expressed as a physical mat-

ter—whether it appears in a paperback edition with a serif

typeface, a hardcover edition with a sans serif typeface, or as

the voice of a narrator on an audio CD. Likewise, the object

code for a software program retains its distinct identity no

'? Sometimes a program's object code is further combined with other

intangible components. For example, an encryption algorithm may be

used to cloak the content of digital transmissions over the public Internet

by changing some 1s to 0s and some 0s to Is in a complex pattern that is

known only to the sending and receiving parties. At the receiving end of

the transmission, the cloak is removed by applying the same algorithm in

reverse. See, e.g., White, supra, at 394-395. Network engineers describe

this technique as operating on a different layer from the application pro-

gram and enclosing the underlying object code of the transmission in a

sealed “envelope” that only the recipient can open. E.g., R. Oppliger, Se-

curity Technologies for the World Wide Web 103-104 (2d ed. 2003); J.

Mairs, VPNs: A Beginner’s Guide 4-6, 9 (2002). As amici Intellectual

Property Professors observe (Br. 5), encryption thereby renders the un-

derlying code “useless gibberish” to third parties who lack the means to

open the envelope. But encryption does not alter the fact that Microsoft

“supplied” the Windows object code to foreign equipment manufacturers.

That is why the object code successfully appeared in unencrypted form on

millions of foreign computers—and why Microsoft collected a licensing fee

each time.

»? See, e.g., Eolas, 399 F.3d at 1339 (“Without this aspect of the pat-

ented invention, the invention would not work at all and thus would not

even qualify as new and ‘useful.””).

27

matter how it is expressed at the physical layer. Just as the

intangible words of Moby-Dick are properly described as a

(particularly essential) component of any audio CD or

printed book bearing that title, object code designed to im-

plement AT&T’s patented invention is a particularly essen-

tial component of that invention, no matter what physical

medium is used to contain or carry it. The same is true of

Microsoft’s own analogy to the player piano (see Br. 21-22,

41): the intangible arrangement of musical notes in The

Star-Spangled Banner is a component—indeed the critical

component—of a player piano configured to play the national

anthem, regardless of whether the arrangement is conveyed

to the piano via a perforated roll or some other physical-

layer technology, such as an IBM punch card, a machine-

readable CD, or a hard drive.”

Citing Pellegrini v. Analog Devices, Inc., 375 F.3d 1113,

1115 (Fed. Cir.), cert. denied, 543 U.S. 1003 (2004), Microsoft

and its amici further claim that intangible object code cannot

be a component of AT&T’s invention because it “is design

information, analogous to product specifications, or a recipe.”

Pet. Br. 38; see also U.S. Br. 14. This argument, too, is base-

less.

In many contexts, a patented device and the instruc-

tions for building it are fully distinct, and it would make no

sense to describe the latter as “components” of the former,

because they are not present in the device. Step-by-step

instructions for building integrated circuit chips (cf. Pelle-

grini, 375 F.3d at 1115) are not themselves part of the fin-

ished product, nor are cookbooks edible. Cf. Pet. Br. 43. But

there is no such distinction between “instructions” and

*' Of course, no one could patent a book or audio CD of Moby-Dick or

an old player-piano capable of playing The Star-Spangled Banner, nor are

the underlying works even entitled to copyright protection in 2007. We

cite these examples simply to demonstrate that, regardless of intellectual

property protection, ordinary speakers of the English language are per-

fectly capable of identifying the separate tangible and intangible “compo-

nents” of a larger system.

28

“product” in the software context, because the product is a

machine that contains and continuously performs the “in-

structions” expressed in object code. Here, the patented

invention is practiced by a computer containing a set of im-

mensely complex instructions for encoding and decoding a

potentially infinite va~iety of voice signals as they arise un-

predictably in real time. The instructions prescribed by the

object code—instructions to open and close circuits in ever-

changing configurations depending on the input—are at all

times present within the computer, whether in storage or

RAM. See Eolas, 399 F. 3d at 1339 (software code is “incor-

porated as an operating element of the ultimate device”); see

also Houston Intellectual Property Law Ass’n Br. 7 (“soft-

ware is capable of becoming incorporated (7.e., stored) within

a patented product”); Yahoo! Br. 10 (“[iJnstallation makes

the software a part of the computer”). It is thus entirely

natural to describe such object code as a component of the

resulting device. -

Il. MICROSOFT “SUPPLIED” INTANGIBLE OBJECT CODE FROM

THE UNITED STATES FOR COMBINATION WITH OTHER COM-

PONENTS ABROAD

A. Mic »soft’s Arguments On The “Supply” Issue Pre-

suppose The Validity Of Its Erroneous Argument On

The “Component” Issue

Once the first question in the petition is answered by

giving the term “component” its ordinary meaning, the sec-

ond question presented is straightforward: Did Microsoft

“suppl[y]” the intangible Windows object code to foreign

computer manufacturers for “combination” with physical

components into devices that, if made in the United States,

would infringe AT&T’s patent? There is no more natural

way to describe what Microsoft has done. To “supply”

means to “satisfy a need or desire for” or to “provide or fur-

nish with.” Webster’s Third Int'l 2297 (citing, as an example

of usage, “a youngster in school supplied me the answer”).

Here, the Windows object code is present in the foreign-

made computers only because Microsoft “provided” or “fur-

nished”—in a word, supplied—it from the United States, via

golden master disk or electronic transmission. Thus, if a

29

shareholder or reporter asked a Microsoft representative

whether the company supplied the codec software that for-

eign Windows users have on their computers, he would an-

swer yes, because that is precisely what Microsoft did. The

same answer follows for purposes of interpreting the words

of this statute.”

Microsoft’s contrary arguments assume that the first

question presented in the petition (and the last question ad-

dressed in Microsoft’s brief) has been answered in Micro-

soft’s favor: i.e., that “component” means “physical compo-

nent” (here, a master disk rather than the code it contains).

The same is true of the Solicitor General, who claims: “The

‘it’ that petitioner supplied from the United States is not the

same ‘it’ that is physically present in any of the foreign-

made computers at issue, i.e., is not a component within the

meaning of the statute.” U.S. Br. 19 (emphasis added). As

Microsoft does throughout its analysis, the Solicitor General

is here assuming that only physical things can be compo-

nents of an invention. That assumption is wrong for the rea-

sons discussed in Point I above.

” The use of the term “supplies” in this context is so obviously ap-

propriate that Microsoft itself conceded in the district court that, when it

sends its Windows object code abroad via electronic transmission, it “sup-

plies its Windows operating system object code from the United States to

certain foreign OEMs.” Pet. App. 46a ¢ 7 (emphasis added); see also U.S.

Br. 4 (stating that Microsoft “provides the Windows object code to foreign

computer manufacturers”). Microsoft has likewise referred to software

companies as “supplier{s]” of software in other proceedings. See United

States v. Microsoft Corp., 253 F.3d 34, 75 (D.C. Cir. 2001) (quoting Micro-

soft’s proposed findings of fact as stating that Symantec is “the leading

supplier of utilities such as anti-virus software’); see also American

Trim, L.L.C. v. Oracle Corp., 383 F.3d 462, 466 (6th Cir. 2004) (“Oracle is a

supplier of business software.”); Dresser-Rand Co. v. Virtual Automa-

tion, Inc., 361 F.3d 831, 837 (5th Cir. 2004) (referring to “the negotiation of

supply agreements for the hardware and software components that were

to make up the control system” (emphasis added)); Specht v. Netscape

Communs. Corp., 306 F.3d 17, 34 n.17 (2d Cir. 2002) (noting that the Uni-

form Computer Information Transactions Act provides guidelines for

“internet-type’ transactions involving the supply of software”); Professor

Edward Lee Br. 6.

30

Microsoft similarly obscures the issue by attempting to

differentiate between the “copy” of the object code used in

personal computers to practice AT&T’s invention and the

“copy” that Microsoft supplied from the United States. But

the word “copy” appears nowhere in Section 271(f); instead,

that provision asks only whether a U.S. company supplied a

“component” from the United States. If the component is

non-physical, as object code is, the term “copy” could have

significance only in describing the different physical-layer

media employed for storing, transporting, or using that

component. There may be many such media, but the object-

code component remains the same. Thus, a foreign pur-

chaser of a personal computer recognizes that the Windows

software inside is the same Microsoft program his neighbor

uses; that it is a core component of his computer; and that it

was created in and supplied from the United States.

Microsoft invokes the Copyright Act in a misguided ef-

fort to justify its focus on “copies” (Br. 19 n.4), but a simple

comparison of the two statutes in fact undermines Micro-

soft’s position. Congress did refer expressly to “copies” in

the Copyright Act and defined them as “material objects” in

which an intangible “work” is “fixed.” 17 U.S.C. § 101. Had

Congress wished to limit Section 271(f) to situations where

the defendant supplied the same “material objects” that end

up in an infringing device abroad, it would have used lan-

guage to that effect. It is noteworthy that Microsoft and its

allies lobbied Congress to take the functional equivalent of

that step by specially limiting the term “component” in Sec-

tion 271(f) to “a tangible item that is itself combined physi-

cally with other components to create the combination that

is alleged to infringe.” Subcomm. on Courts, the Internet,

and Intellectual Property of the House Comm. on the Judici-

ary, Committee Print: Patent Act of 2005, § 10, at 49 (Apr.

14, 2005) (proposing new section 271(f)(3)). But if such a

limitation were warranted as a policy matter, it should be

imposed by Congress, not by this Court.

Microsoft also contends (Br. 42) that, before object code

“can be ‘supplied,’ . . . it first must be reduced to some physi-

31

cal format,” and that object code unencased in such a physi-

cal format “is not susceptible to transmission.” That is true,

but it does not support Microsoft’s position. Of course Mi-

crosoft can supply object code to foreign manufacturers only

if it first encases the code in a physical-layer container, but

the object code itself remains the component supplied. The

fact that this component can be moved seamlessly from one

container (such as a master disk) to other containers (such as

a magnetic surface on a computer hard drive or electrical

charges on a RAM chip) merely illustrates the ease with

which Microsoft supplies that component from the United

States for combination with other components abroad “in a

manner that would infringe [AT&T’s] patent if such combi-

nation occurred within the United States.” 35 U.S.C.

§ 271(f).

By analogy, suppose that a foreign publisher wishes to

print and sel] ten thousand copies of Moby-Dick in its home

country. It can supply the ink, paper, and printing presses,

but it lacks the most important component of the book: the

complete and accurate sequence of Melville’s words from the

beginning of the novel to the end. It therefore contracts

with an American company to convey that word sequence

via an electronic transmission. The foreign company

downloads the word sequence, chooses a typeface and page

format, prints the book, and sells it to consumers. In ordi-

nary speech, the American company has plainly “supplied,”

from the United States, the intangible text of Moby-Dick for

combination with the physical components of the printed

books, even though the physical format of Moby-Dick as it

appears in the books is obviously different from the elec-

tronic format in which the American firm transmitted it

abroad. Likewise, Microsoft has also “supplied,” from the

United States, intangible object code for combination with

various physical components to produce devices abroad that,

if manufactured here would infringe AT&T's patent.”’

** As before, we cite this example simply to show how the terms

“component” and “supply” are ordinarily used in analogous contexts, not

32

Microsoft claims (Br. 42) that “it is impossible to deter-

mine the location from which [software] is supplied” if that

software is conceptualized as intangible object code rather

than a physical medium containing that code. This is sophis-

try. Microsoft concedes that it “conceived, wrote, compiled,

tested, and debugged Windows in the United States” (Mi-

crosoft Ct. App. Br. 4; see also U.S. Br. 4) and then shipped

the Windows software abroad by combining it with a physi-

cal-layer medium in the United States. Indeed, Microsoft

had no difficulty acknowledging the U.S. origin of Windows

software installed on foreign-made computers when doing so

allowed Microsoft to reap $31 million in tax deductions. Mi-

crosoft stipulated that its “software development in the

United States” satisfied the “domestic production require-

ment” necessary for master disks containing Windows to be

treated as deductible “export property” under 26 U.S.C.

§ 927(a)(2)(B) (repealed 2000). Microsoft Corp. v. Commis-

sioner, 311 F.3d 1178, 1182 (9th Cir. 2002). Microsoft’s tax

deduction for “export property” applied not only to the

physical disks exported from the United States, but to all

“royalties that Microsoft earned” from licenses to foreign

OEMs, including the “royalty for each copy of the [software]

distributed in the market or for each computer system the

OEMs sold.” /d. at 1181.

There also can be no doubt that the intangible object

code meets the other statutory criteria for liability under

both paragraphs (1) and (2) of Section 271(f). Microsoft has

never denied that, if object code itself is a component of

AT&T’s invention, it constitutes “a substantial portion of

the components” of that invention for purposes of paragraph

to demonstrate anything about how intellectual property law might treat

this transmission of Moby-Dick, which has long been in the public domain.

Although Microsoft notes (Br. 26 n.6) that software may be protected by

copyright law, AT&T’s invention consists of “new and useful” methods

and products and is therefore protected from misappropriation only by

patent law, not by copyright. Microsoft wrote the object code (which

could be protected by copyright) that, when combined with physical com-

puter components, infringed AT&T's patent.

33

(1). 35 U.S.C. § 271(f)(1) (emphasis added). That alone is

enough to establish Section 271(f) liability because it is un-

disputed that, through its contracts with foreign equipment

manufacturers, Microsoft “actively induce{d]” the installa-

tion of this object code in computers “outside of the United

States in a manner that would infringe the patent if such

combination occurred within the United States.” Jd.

Microsoft is independently liable under paragraph (2) as

well. The speech-codec object code included with the Win-

dows operating system is “especially made or especially

adapted for use in” AT&T’s invention; it is “not a staple arti-

cle or commodity of commerce suitable for substantial nonin-

fringing use”; and Microsoft transmitted it abroad “knowing

that such component is so made or adapted and intending

that such component wiould] be combined outside of the

United States in a manner that would infringe the patent if

such combination occurred within the United States.” 35

U.S.C. § 271(f)(2). Microsoft contends that it cannot be liable

under paragraph (2) because “Windows is not. ‘especially

made or especially adapted for use” in AT&T’s invention.

Pet. Br. 12 n.3 (emphasis added). This is untenable. The

question is whether the portion of the Windows object code

relevant to speech encoding and decoding, not Windows as a

whole, is “especially made or especially adapted for use” in

AT&T’s invention. It indisputably is, because that portion of

the code has no purpose other than the coding and decoding

of speech signals.

Finally, it is difficult to discern what Microsoft hopes to

gain by arguing that, if it is liable for violating Section 271(f),

it is liable, “at most, for a single act of infringement for each

master version shipped overseas.” Pet. Br. 24. The lower

courts held only that Microsoft was liable under Section

271(f), and that liability determination is the only question

presented for this Court’s review. Microsoft appears to be

asking the Court to opine on the measure of the damages

appropriate to compensate AT&T for that violation, but the

lower courts never undertook to ascertain damages in this

case, and no damages issue is before this Court. Indeed, the

34

parties have entered into a settlement agreement that pre-

scribes fixed dollar outcomes depending on the course of ap-

pellate proceedings about the underlying question of liabil-

ity. See Pet. 9, Pet. App. 42a-43a.

In any event, Microsoft cannot seriously complain about

the fairness of making AT&T whole “for each of the tens of

millions of foreign-produced copies” (Br. 24). Microsoft itself

encouraged its foreign business partners to sell Windows to

as many end users as possible, kept close track of how many

times they did so, and charged them a royalty each time. It

would hardly be unreasonable to take account of Microsoft’s

royalties in calculating its liability for its intentional in-

fringement of AT&T’s patent. See 35 U.S.C. § 284 (in-

fringement creates right to “damages adequate to compen-

sate for the infringement”); General Motors Corp. v. Devex

Corp., 461 U.S. 648, 654 (1983) (successful plaintiff is entitled

to “full compensation for ‘any damages’ he suffered as a re-

sult of the infringement” (citation omitted)); see also Deere &

Co. v. International Harvester Co., 710 F.2d 1551, 1558-1559

(Fed. Cir. 1983); cf. Professor Edward Lee Br. 13 (“a court

should be allowed to consider, when computing damages for

profits lost by AT&T, the acts of copying by Microsoft’s li-

censees that Microsoft’s predicate act of infringement facili-

tated”).”*

** Quite apart from all of the considerations discussed to this point,

Microsoft would be liable under Section 271(f) even if, as Microsoft erro-

neously claims, it could infringe only by supplying a physical component

abroad. Section 271(f) makes it an act of infringement to supply “compo-

nents” abroad “in such manner as to actively induce the combination of

such components outside of the United States.” 35 U.S.C. § 271(f)(1) (em-

phasis added). Microsoft's position depends not just on its artificially nar-

row. construction of “component,” but also on the premise that “such,” as

used in this sentence, means “the same” down to the last molecule. E.¢.,

Pet. Br. 8, 14, 15, 18. But the word “such” cannot bear that weight. Even

under the narrowest definition of that word-—“aforementioned”— it de-

notes only substantial, not literal, identity between two things. One can

satisfy instructions to “buy three eggs and combine such eggs with flour”

even if one removes the shells in the process. Here, as with eggs, the es-

sence of software is not the shell it comes in, but the contents, which

35

B. Microsoft's Invocations Of Legislative History Are

Unavailing

As discussed, the ordinary meaning of the statutory

text answers both questions in this case. Microsoft nonethe-

less argues that Section 271(f) should be construed narrowly

in light of its “legislative history.” Pet. Br. 36 n.10; see also

U.S. Br. 17. Congress, it says, enacted Section 271(f) only

because it “was concerned with the specific facts of Deep-

south” (Pet. Br. 36 n.10), where a defendant avoided patent

liability even though it had supplied all of the physical “com-

ponents” of a shrimp deveining machine for assembly

abroad. But while Deepsouth was indeed the catalyst for the

enactment of Section 271(f), “statutory proh‘bitions often go

beyond the principal evil to cover reasonably comparable

evils, and it is ultimately the provisions of our laws rather

than the principal concerns of our legislators »y which we

are governed.” Oncale v. Sundowner Offshore Servs., Inc.,

523 U.S. 75, 79 (1998). As noted, software was well-

established as a basis for patentable inventions when Con-

gress enacted Section 271(f) in 1984. See supra pp. 22-23 &

n.16. If Congress had meant to restrict the scope of Section

271(f) to physical components like those at issue in Deep-

south, thereby denying meaningful protection to any inven-

tion that could be practiced by the use of software, it would

have said so in the provision’s text.

Indeed, Section 271(f) indisputably creates patent liabil-

ity for a range of conduct beyond the type of activity ad-

dressed in Deepsouth. In that case, the defendant supplied

all of the components of the patented shrimp deveining ma-

chine for combination abroad. See Deepsouth, 406 U.S. at

524. But a defendant can be liable under Section 271(f)(1) for

supplying only “a substantial portion” of the components of a

patented invention. A defendant can also be liable under

Section 271(f)(2) for supplying a single “component” of that

invention if (as in this case) the component is especially

clearly are transferred onto (and thus “combined with”) the foreign-made

computers.

36

made for use in the invention, and liability attaches even if

that one component is never actually combined abroad, so

long as the defendant intends that such a combination take

place. See Waymark Corp. v. Porta Sys. Corp., 245 F.3d

1364, 1367-1368 (Fed. Cir. 2001). Because Section 271(f) thus

extends far beyond the facts of Deepsouth, that case cannot

provide a basis for carving arbitrary exceptions out of the

statutory language.

III. MICROSOFT IDENTIFIES NO POLicy BASIS FoR DISREGARD-

ING THE PLAIN MEANING OF THE STATUTORY TEXT

A. Far From Vindicating The “Technology-Neutral”

Purposes Of Section 271(f), Microsoft’s Position

Would Repeal That Provision For Software Compo-

nents

Microsoft and its supporters argue that the court of ap-

peals’ holding offends the “technology-neutral” (U.S. Br. 25)

objectives of the statutory scheme by precluding software

companies from conducting research and development ac-

tivities in the United States while exploiting the fruits of

those activities abroad through foreign manufacturing op-

erations. In fact, principles of technological neutrality cut

against Microsoft here, because it is Microsoft’s position that

would single out software for special treatment by arbitrar-

ily exempting it from the scope of Section 271(f).

As an initial matter, the court of appeals’ holding does

not preclude software designers from conducting research

and development in the United States as one step towards

manufacturing abroad a product that would violate U.S. pat-

ents if manufactured here. For example, nothing in the

court of appeals’ decision precludes a company in Microsoft’s

position from providing foreign companies with technical

specifications—the software equivalent of blueprints—for

particular types of software programs.”> What Microsoft

*5 Once the “requirements” of a desired program—i.e., what the pro-

gram should do—have been established, software engineers create techni-

cal specifications (often called the “design”), the purpose of which is “to

create a model of software that will implement all customer requirements

37

may not do is what it did here: supply the actual object code

that will itself be combined with hardware to create devices

that would infringe patents if manufactured in the United

States. The Solicitor General contends that it would “upset

the balance struck by Congress” to encourage software

companies like Microsoft, which balk at paying royalties to

inventors, to “sell[{] incomplete work product” abroad in the

form of design specifications rather than ready-to-install ob-

ject code. U.S. Br. 26. But permitting such companies to

free-ride on others’ inventions by selling complete work

product for the most critical component in those inven-

tions—finished object code—would strike no balance at all:

it would repeal Section 271(f) for the software industry.

Indeed, a repeal is precisely what Microsoft and its al-

lies seek. First, they are actively lobbying for legislation

that, in its most aggressive form, would eliminate Section

271(f) outright. See Patent Reform Act of 2006, S. 3818,

109th Cong. § 5(f) (2006); Eli Lilly Br. 14-15 (stating that Mi-

crosoft’s coalition is pursuing “sweeping changes to U.S.

patent law,” including a repeal of Section 271(f)). Second, as

a hedge, they have asked this Court to issue the functional

equivalent of a repeal for the software industry in this case.

correctly.” Pressman, supra, at 250. Design specifications are not them-

selves software, but rather representations of the software’s structure:

“the preliminary blueprint from which software is constructed,” id. at 254.

Completion of the design specification “sets the stage for construction,”

which is the actual generation of source code. /d. at 227; see also id. at 321

(stating that the later phase of software design, called “component-level

design,” represents the program “in sufficient detail to guide in the gen-

eration of programming language source code”). Once generated, the

source code must be compiled into machine-executable object code and

actually run on a computer for testing and “debugging” (removal of errors

identified during testing). See Computer Assocs. Int'l, Inc. v. Altai, Inc.,

982 F.2d 693, 698 (2d Cir. 1992). As the court of appeals correctly held,

this case does not involve the transmission overseas of software “designs”

or specifications that would instruct foreign engineers on how to code

Windows themselves. See Pet. App. 8a. Rather, Microsoft shipped the

final software component—the program’s object code—for incorporation

into foreign-made computers.

38

On this the Court should make no mistake: the consequence

of Microsoft’s position is that, with few if any exceptions, no

software “component” could ever be “supplied” from the

United States in a manner that would trigger Section 271(f).

As Microsoft acknowledges (albeit in a footnote), every

provision of object code to an end user necessarily involves

transferring that code from one physical-layer container to

another. See Pet. Br. 4 n.2 (conceding that “the ‘installation’

process itself involves an act of duplication”). If Microsoft

were correct that only the physical medium containing the

object code qualifies as a component “supplied” from the

United States—such that the “supply” of object code is in-

terrupted whenever the code is transferred to a new physi-

cal medium—the installation process would almost always

generate a new “copy” that (under Microsoft’s theory) is not

“supplied” from the United States. Microsoft could avoid

liability even if it directly mailed each individual foreign cus-

tomer a CD containing all of the relevant object code, be-

cause each customer would normally install the object code

onto a hard drive and, in the process, transfer the object

code from one physical medium (the pits and lands of a port-

able CD) to another (the magnetic surface on hard drives).

In the district court, Microsoft tried to avoid that con-

clusion by “acknowledg{ing] that if individual disks with the

infringing Windows object code were sent abroad for instal-

lation into each foreign-assembled computer (rather than

one golden master disk), Microsoft would be liable for in-

fringement under Section 271(f).” Pet. App. 36a n.7; J.A. 26.

And the Solicitor General makes the same assertion even

now (Br. 25 n.2). But these attempts to seem moderate run

headlong into Microsoft’s own core legal rationale, which is

that every act of “copying” (“installation”) outside the

United States creates a new software component that was

not “supplied” from the United States. That is presumably

why Microsoft has now retreated from any suggestion that it

would be liable if it had sent millions of CDs containing the

object code needed to practice AT&T’s invention to millions

39

of foreign end users for downloading onto their individual

hard drives.”

Microsoft nonetheless claims that Section 271(f) “might”

preclude a company in its position from shipping “software-

encoded disks” to foreign manufacturers for physical incor-

poration into individual computers (Pet. Br. 28), but only in

the atypical case (such as certain “video game systems”)

where the end user must “run the computer program di-

rectly from the U.S.-supplied disk” rather than downloading

it onto the computer’s hard drive (id. at 37 n.11). Even on its

own terms, this effort to preserve relevance for Section

271(f) in the software industry has vanishingly narrow sig-

nificance. End users typically download programs from a

disk onto a hard drive before running them, and Microsoft

(and similar companies) could easily ensure that end users

follow that step for all programs if that were all that is

needed to avoid liability under Section 271(f).

In addition, under the logic of Microsoft’s argument, a

company in Microsoft’s position would not “supply” foreign

end users with the same “copy” of the object code actually

employed to practice AT&T’s invention even if Microsoft

sent all of those end users the hard drives themselves, fully

equipped with the code. That is because, before a computer

can perform the operations of a software program, it must

call up the object code from a storage medium (such as a

hard drive or CD) and incorporate it into arrangements of

© The Solicitor General, on the other hand, continues to argue that if

Microsoft had “sent copies of its Windows software from the United

States to a foreign country and those copies were loaded onto computers,

{it} would likely be liable under Section 271(f) for each such infringing

copy.” U.S. Br. 25 n.2. Again, however, “loading” software from a CD

onto a computer’s hard drive inevitably requires converting code from its

physical manifestation as pits and lands on the CD into a new physical

manifestation as magnetic properties on the hard drive (cf. id. at 15). The

Solicitor General offers no principled reason for treating the different

physical manifestations of object code on a CD and a hard drive as the

same component while treating the different physical manifestations of

object code on the master disk and a hard drive as different components.

40

electrical charges in its RAM.”’ Only once the code is so em-

bodied in the RAM circuitry can the computer’s central

processor perform the program’s functions (here, speech en-

coding and decoding operations). On Microsoft’s theory, the

“copy” of the Windows object code embodied in the com-

puter’s RAM would be different from the “copy” physically

embodied in the storage medium. For that matter, even the

electric charges in the RAM circuitry discharge over time

and must be “refreshed”—by adding new electrons that are

different from those previously discharged—“thousands of

times per second.””*

One way or another, Microsoft’s legal rationale would al-

low it to claim that, no matter how it distributes software

abroad, it never “supplies” the same software component

that foreign customers actually use to practice AT&T’s in-

vention—even though there is obviously no other way for

this staggeringly complex compilation of object code to end

up in the RAM of millions of foreign customers’ computers.

Microsoft’s angels-on-a-pin metaphysics is reminiscent of the

claim attributed to Heraclitus that “you would not step

twice into the same river” because other waters are continu-

ally flowing in.” But ordinary speakers of the English lan-

guage find it perfectly natural to say that Huck and Jim

*” See Walters, supra, at 41 (“Programs that are executing—actually

in use—are loaded into [RAM] because the silicon chips that comprise

[RAM] can read and store data much faster than can the other principal

kind of storage, hard disks.”); Jeff Tyson, How Computer Memory Works,

at http://computer.howstuffworks.com/computer-memory.htm (visited

Jan. 23, 2007).

** Jeff Tyson & Dave Coustan, How RAM Works, at http://computer

-howstuffworks.com/ram.htm (visited Jan. 23, 2007); see also White, supra,

at 49. Under Microsoft’s position, therefore, the functional software

“component” is never the same from one moment to the next even while it

is being used. See, e.g., Apple Computer, Inc. v. Formula Int'l, Inc., 594

F. Supp. 617, 622 (C.D. Cal. 1984) (“It is a property of RAM that when the

computer is turned off, the copy of the programs recorded in RAM is

lost.”); White, supra, at 49.

* Plato, Cratylus 402a, quoted in G.S. Kirk & J.E. Raven, The

Presocratic Philosophers 197 n.218 (1971).

41

rafted on the same Mississippi River on successive days, de-

spite the differences in water molecules. So, too, do ordinary

speakers find it perfectly natural to say that Microsoft “sup-

plied” the Windows object code to its foreign customers

abroad, and that the foreign customers sold computers that

run “Microsoft Windows” software made in the United

States, not “Sony Windows” made in Japan or “Siemens

Windows” made in Germany.

In sum, Microsoft’s position would treat the software

industry differently from all other industries by precluding

any meaningful application—indeed, any application at all—

of Section 271(f) to the supply of software components

abroad. It is Microsoft and its allies, not the court of ap-

peals, that would thereby thwart principles of “technological

neutrality.” And there is absolutely no indication that Con-

gress meant to treat software any differently from any other

components of patented inventions, much less that it in-

tended to fence software off from the protections of Section

271(f) entirely.

B. The Presumption Against Extraterritoriality Is In-

applicable

In another effort to escape the statutory text, Microsoft

relies heavily on the judicial “presumption against the extra-

territorial application of U.S. law” (Pet. Br. 30). That reli-

ance is misplaced for multiple independent reasons.

First, the presumption is simply inapplicable because

Microsoft is liable under Section 271(f) not for “extraterrito-

rial” conduct, but for conduct performed domestically: ship-

ping its U.S.-developed, U.S.-tested Windows software from

the United States. Section 271(f)(2) makes this point abun-

dantly clear. Because that provision predicates liability on

the “inten(t]” of a U.S. firm in shipping a “component”

abroad, Microsoft is liable no matter what actually happened

abroad and, in particular, whether or not foreign manufac-

turers ultimately combined that component into devices that

would infringe AT&T’s patent if made in the United States.

See Waymark Corp., 245 F.3d at 1367-1368. As noted, the

proper assessment of damages may well turn on the extent

42

to which Microsoft “actively induce{d],” 35 U.S.C. § 271(f)(1),

or “intend[ed],” id. § 271(f)(2), the proliferation of infringing

devices abroad. But even proof on that damages issue would

require no foreign discovery, since Microsoft kept track of

its foreign partners’ activities and charged them royalties

that it entered into its books in the United States. Holding

Microsoft liable under that provision no more entails the

“extraterritorial” application of the patent law than would

holding a criminal defendant liable under U.S. law for mail-

ing an explosive device from the United States with the in-

tent to harm a target abroad. See 18 U.S.C. § 1716(j)(2).

Second, even if Microsoft’s activity in this case had an

extraterritorial dimension, the “presumption” would be in-

applicable because, where it applies at all, it operates only to

break interpretive ties when a statute is ambiguous as to its

geographic scope. See, e.g., EEOC v. Arabian Am. Oil Co.,

499 U.S. 244, 248 (1991) (stating that the presumption

against extraterritoriality applies when ascertaining “‘unex-

pressed congressional intent” (quoting Foley Bros., Inc. v.

Filardo, 336 U.S. 281, 285 (1949)). Because, for the reasons

discussed in Points I and II above, the text of Section 271(f)

speaks directly to the questions presented here, there is no

ambiguity for any “presumption” to resolve.

Third, it would be especially perverse to apply the pre-

sumption against extraterritorial application to a statutory

provision that, like this one, Congress enacted for the pur-

pose of overcoming the application of that very presumption.

Congress passed Section 271(f) precisely to counteract the

Deepsouth Court’s reliance on the presumption against ex-

traterritoriality to foreclose the type of liability that Section

271(f) creates.” Thus, the premise of Section 271(f) is that,

© See Deepsouth, 406 U.S. at 531. Nothing in the reasoning of Deep-

south supports Microsoft’s claim that applying Section 271(f) in these cir-

cumstances would involve an “extraterritorial” application of U.S. law

(and it would not, for the reasons discussed in the text). In Deepsouth,

because Section 271(f) had not yet been enacted, the U.S. company could

face no liability for infringing a combination patent unless Section 271(a)

had genuinely extraterritorial application, in the sense that a foreign com-

43

in the circumstances defined by that provision’s plain lan-

guage, U.S. patentees should not have to rely exclusively on

foreign patent protections, which would require the patentee

to assume prohibitive costs to prosecute and enforce patents

in scores of foreign jurisdictions, all simply to address do-

mestic activity by U.S. companies like Microsoft that supply

components of the patented invention from the United

States. Given that backdrop, applying a judge-made “pre-

sumption” to reweigh the policy balance that Congress

struck when writing the language of Section 271(f) would

hardly give effect to any presumed congressional intent; in-

stead, it would usurp Congress’s role. *

The Solicitor General’s contrary argument (see U.S. Br.

28) relies on two cases that do not begin to support it. The

Solicitor General’s reliance on Smith v. United States, 507

U.S. 197 (1993) (U.S. Br. 28), is particularly baffling, since

the statute in that case—the Federal Tort Claims Act—

expressly foreclosed any extraterritorial application. /d. at

201 (“(T]he FTCA’s waiver of sovereign immunity does not

apply to ‘[aJny claim arising in a foreign country.” (quoting

28 U.S.C. § 2680(k))). Thus, this Court held that the pre-

sumption against extraterritoriality was “doubly fortified by

the language of this statute.” Jd. at 204 (quoting United

States v. Spelar, 338 U.S. 217, 222 (1949)). The exact oppo-

site is true of Section 271(f): it was enacted to reverse this

Court’s reliance on the presumption against extraterritorial-

ity in Deepsouth. See 406 U.S. at 531.

pany could violate U.S. patent laws (or a U.S. company could “induce”

such a violation under Section 271(b)) through assembly of an infringing

device abroad. The Court rightly noted that Congress had not revealed

an intention to create liability in those circumstances.

*' In claiming that U.S. patent law is “territorially limited,” Micro-

soft tellingly relies entirely on statements that predate the enactment of

Section 271(f). See Pet. Br. 31 (citing Dowagiac Mfg. Co. v. Minnesota

Moline Plow Co., 235 U.S. 641 (1915); Brown v. Duchesne, © US. (19

How.) 183 (1857)); see also U.S. Br. 27 (same).

44

F. Hoffmann-La Roche Ltd. v. Empagran S.A., 542

U.S. 155 (2004), is likewise inapposite. Far from adopting a

generalized presumption against the extraterritorial appli-

cation of U.S. law, the Court there noted only that “it ordi-

narily construes ambiguous statutes to avoid unreasonable

interference with the sovereign authority of other nations.”

Id. at 164 (emphasis added). It thus rejected an interpreta-

tion of U.S. law that would have vested “worldwide subject

matter jurisdiction” in the U.S. courts for “any foreign suitor

wishing to sue its own local [foreign] supplier” for foreign

injuries, “provided that a different plaintiff had a cause of

action against a different firm for injuries that were within

U.S. ... commerce.” Jd. at 166 (internal quotation marks

omitted). Congress, the Court held, could not have intended

that bizarre result: “Why should American law supplant, for

example, Canada’s or Great Britain’s or Japan’s own deter-

mination about how best to protect Canadian or British or

Japanese customers from anticompetitive conduct engaged

in significant part by Canadian or British or Japanese or

other foreign companies? .... We can find no good answer

to the question.” /d. at 165-166. Here, in contrast, it would

hardly be “unreasonable” to apply Section 271(f) in this suit

brought by one U.S. company against another U.S. company

for actions taken within the United States.

Fourth, applying any presumption against extraterrito-

riality would be particularly pointless from a policy perspec-

tive, because Microsoft cites no respect in which the applica-

tion of Section 271(f) would actually conflict with any foreign

nation’s sovereign prerogative. As Microsoft acknowledges

(Br. 30), the presumption against extraterritoriality “serves

to protect against unintended clashes between our laws and

those of other nations which could result in international dis-

cord.” Arabian Am. Oil, 499 U.S. at 248. Neither in this

Court nor below has Microsoft identified any law of any for-

eign nation that poses any conflict with the application of

Section 271(f) here. Certainly no foreign law requires Micro-

soft to infringe AT&T’s patent by shipping Windows object

code from the United States to foreign computer manufac-

turers without paying AT&T a reasonable royalty for ex-

45

ploiting its invention.” It is also noteworthy that no foreign

government has submitted any objection to the application

of Section 271(f) in this case, whereas several foreign gov-

ernments did file amicus briefs in Empagran urging the re-

sult the Court ultimately reached. See 542 U.S. at 167-168.

And the Solicitor General, who represents the Executive

Branch before this Court, nowhere suggests that this case

implicates the United States’ obligations under international

law or the President’s authority over foreign relations.”

The mere fact that countries have different patent-law

regimes does not create a conflict between those patent re-

* See Hartford Fire Ins. Co. v. California, 509 U.S. 764, 799 (1993)

(“Since [petitioners] do not argue that British law requires them to act in

some fashion prohibited by the law of the United States, . . . or claim that

their compliance with the laws of both countries is otherwise impossible,

we see no conflict with British law.”); cf. Steele v. Bulova Watch Co., 344

U.S. 280, 285-286 (1952) (“{T}he United States is not debarred by any rule

of international law from governing the conduct of its own citizens . . . in

foreign countries when the rights of other nations or their nationals 1 are

not infringed.” (internal quotation marks omitted)). The facts of Steele are

instructive. This Court there held that a U.S. district court could hear a

Lanham Act claim against a trademark infringer whose infringing prod-

ucts were sold only in Mexico—a situation that certainly raises greater

extraterritoriality concerns than holding Microsoft liable for the U.S.-

based development, testing, debugging, and shipping of software. The

Court even contemplated the possibility of a U.S. court enjoining the U.S.

defendant “to cease or perform acts” occurring entirely in Mexico, so long

as there was “no conflict which might afford [the party] a pretext that

such relief would impugn foreign law.” Steele, 344 U.S. at 289. In so hold-

ing, the Court rejected the dissent’s view—echoed by Microsoft here—

that an unspecified and unsubstantiated risk of “conflict with the laws and

practices of other nations” warranted exempting U.S. parties from the

reach of a U.S. statute. /d. at 258-259 (Reed, J., dissenting).

*S If anything, faithful enforcement of Section 271(f) comports with

the international goal of fighting cross-border piracy by ensuring that

countries maintain “expeditious remedies to prevent infringements and

remedies which constitute a deterrent to further infringements.” Agree-

ment on Trade-Related Aspects of Intellectual Property Rights, Including

Trade in Counterfeit Goods, Apr. 15, 1994, art. 41(1), in Marrakesh

Agreement Establishing the World Trade Organization, Annex 1C, 33

I.L.M. 1197, 1213-1214 (1994).

46

gimes and Section 271(f). Section 271(f) merely defines a

category of U.S.-based behavior as infringing of U.S. pat-

ents; it does not purport to give any individual a monopoly

over the making of any invention in a foreign country. Nor

has AT&T ever contended, contrary to Microsoft’s implica-

tion, that it was an “act of infringement” to “assemble de-

vices overseas” (Pet. Br. 33). The foreign computer manu-

facturers that combined the Windo\’s object code inte their

foreign-made computers have nothing to fear from the U.S.

patent law, provided that the resulting devices are not “im-

port(ed] into the United States” (85 U.S.C. § 271{a)). And

Microsoft cites no basis for concern that any foreign court

would hold it liable for the same conduct underlying Micro-

soft’s Section 271(f) violation (much less deny Microsoft the

ordinary right of any defendant to offset damages in a prior

case from damages assessed in a subsequent case). Micro-

soft’s claim of “overlapping and duplicative liability” (Pet.

Br. 31 n.8) is therefore neither substantiated nor realistic.

Finally, there is no merit to Microsoft’s related policy

argument that, to vindicate their rights abroad, patent-

holders in AT&T’s position should simply rely on patent pro-

tections under foreign law rather than on U.S. patent law.

Again, Congress enacted Section 271(f) because it under-

stood that foreign patent protections are sometimes weaker

than their U.S. counterparts, and because it wished to spare

U.S. patent-holders from the considerable expense of obtain-

ing patent protections in dozens of foreign jurisdictions. Mi-

crosoft’s policy argument could be made whenever Section

271(f) is invoked, because that provision necessarily entitles

U.S. patent-holders to seek redress for U.S.-based actions as

an alternative to seeking redress for foreign acts under for-

eign patent-law protections. The argument thus quarrels

not so much with the application of Section 271(f) in this case

as with Congress’s decision to enact it in the first place.”

- Judge Learned Hand, faced with the question whether a copyright

infringer was liable for the shipping of film negatives abroad from which a

copyrighted film could be reproduced, noted that it was irrelevant

47

C. Microsoft's “Outsourcing” Arguments Are Without

Merit

Microsoft and its supporters argue (e.g., Pet. 20; U.S.

Br. 26; Business Software Alliance Br. 10, 12; Intel Br. 19)

that the Court should limit the scope of Section 271(f) on the

theory that construing it as the court of appeals did would

lead companies in Microsoft’s position to relocate their op-

erations abroad. But even if policy arguments could trump

the plain meaning of statutory text, which they cannot, this

policy argument is unpersuasive on the merits.

As an initial matter, arguments about supposed “out-

sourcing” incentives, like Microsoft’s other policy-based ar-

guments, take issue less with the court of appeals’ decision

than with Congress’s threshold decision to enact Section

271(f). No matter how it is interpreted, that provision has

always presented the hypothetical risk that “component”

suppliers will move their operations offshore. See D. Chi-

sum, Normative and Empirical Territoriality in Intellec-

tual Property: Lessons from Patent Law, 37 Va. J. Int'l L.

603, 607 (1997) (calling for repeal of Section 271(f)). Con-

gress nonetheless decided to protect U.S. patentees from

unfair competition by U.S. “component” suppliers. To the

extent that faithful application of Section 271(f) may some-

day have undesirable economic consequences, Congress is

more than capable of fine-tuning it to strike a different bal-

ance between inventors and those who wish to free-ride on

others’ innovations. While Congress has considered such

fine-tuning, it has not changed the law yet.”

whether the copyright holder would be entitled to recover under foreign

law. See Sheldon v. Metro-Goldwyn Pictures Corp., 106 F.2d 45, 52 (2d

Cir. 1939), affd, 309 U.S. 390 (1940). Although Sheldon arose under the

copyright law, not the patent law, it supports the court of appeals’ conclu-

sion here that the consequences under U.S. law for the U.S.-based actions

of U.S. companies do not turn on foreign law. Pet. App. 6a n.2.

*° Microsoft’s argument on this point also ignores the undisputed

ability of any U.S. company, under any interpretation of Section 271(f), to

supply work product (such as technical specifications) from earlier stages

of the soft ware-development process. See supra p. 36 & n.25.

48

Microsoft’s policy argument also focuses narrowly on

the welfare of U.S.-based component suppliers, a myopic

perspective that—though favorable to Microsoft and its al-

lies—slights the fuller set of policy interests Congress

sought to accommodate by enacting this statute. Section

271(f) encourages greater innovation in the U.S. by ensuring

effective intellectual property protection for inventors, and

all software developers (including Microsoft) benefit from

that protection. As one senior U.S. executive explained:

The software industry could achieve cost savings by

moving its development and production facilities

overseas, but it has chosen to remain in the United

States and has flourished here, in no small part be-

cause the copyright, trade secret and judicial proc-

esses in the United States provide[] strong and ef-

fective protection for the intellectual property con-

tent of software products. There is no justification

for letting them enjoy the benefit of our strong IP

system for their own products while, at the same

time, they are allowed to avoid exposure to other

companies’ patents when those same products are

exported.”

*° Committee Print Regarding Patent Quality Improvement: Hear-

ing Before the Subcomm. on Courts, the Internet, and Intellectual Prop-

erty of the House Comm. on the Judiciary, 109th Cong., Ist Sess. 202

(2005) (statement of Jack Haken, Vice President, Intellectual Property &

Standards, U.S. Philips Corporation). Amicus Business Software Alliance

acknowledges (Br. 9) that “[sJoftware and computer companies based in

the United States rely on the strength of United States [patent] law” to

protect their own discoveries, yet claims (Br. 10) that Section 271(f) dis-

advantages U.S. software companies compared to foreign companies. But

there is no disparity in treatment; a foreign company can no more use the

U.S. market to develop “components” of a U.S.-patented invention for

combination abroad than can a U.S. company. And U.S. companies may

set up manufacturing facilities in other countries just as foreign companies

do, provided that they do not ship “components” of U.S.-patented inven-

tions for combination with other components—or, if they do, that they

license the technology from the patentee, like any other U.S. company

seeking to exploit a U.S. patent.

49

The lower courts’ approach also removes the artificial

incentives that Microsoft’s position would give to end-

product manufacturers {in this particular case, computer

manufacturers) to locate their operations offshore in order to

exploit Microsoft’s proposed software exception to Section

271(f). Under Microsoft’s position, any manufacturer that

depends on U.S.-developed software gains substantial sav-

ings (i.e., avoidance of patent royalties) by locating its manu-

facturing facilities abroad. A proper construction of Section

271(f) levels the playing field among jurisdictions by ensur-

ing that patent-related costs will be the same wherever the

manufacture occurs. See 130 Cong. Rec. 28,069 (1984), re-

printed in 1984 U.S.C.C.A.N. 5827 (stating that Section

271(f) was enacted “to avoid encouraging manufacturing

outside the United States”).”” As a result, U.S. business

leaders who oppose a repeal of Section 271(f) “passionately

argue that Section 271(f) protects [certain] kinds of jobs

(such as assembly jobs) from foreign outsourcing.” Amend-

ment in the Nature of a Substitute to H.R. 2795, the “Patent

Act of 2005”: Hearing Before the Subcomm. on Courts, the

Internet, and Intellectual Property of the House Comm. on

the Judiciary, 109th Cong., Ist Sess. 68 (2005) (statement of

Phil Johnson, Chief Patent Counsel, Johnson & Johnson).

There is therefore “a sharp difference of opinion within

the IP stakeholder community as to whether, in the future,

Section 271(f) will result in a net increase or decrease in U.S.

jobs.” Jd. More generally, there is no consensus about the

net effect of Section 271(f) on U.S. economic interests. The

responsibility for resolving that empirical controversy rests

with Congress, not this Court.

The Solicitor General incorrectly asserts (Br. 27 n.3) that this

statement applied only to a “different provision” of the same bill. Con-

gress was clear that the bill contained “two major changes in the patent

law to avoid encouraging manufacturing outside the United States,” the

“second” of which was Section 271(f). 130 Cong. Rec. 28,069, reprinted in

1984 U.S.C.C.A.N. 5827-5828 (emphasis added); see also Bayer AG v.

Housey Pharms., Inc., 3A0 F.3d 1367, 1374 (Fed. Cir. 2003).

50

CONCLUSION _

The judgment of the court of appeals should be affirmed.

Respectfully submitted.

LAWRENCE J. LAFARO SETH P. WAXMAN

THOMAS A. RESTAINO Counsel of Record

AT&T CORP. WILLIAM G. MCELWAIN

One AT&T Way JONATHAN E. NUECHTERLEIN

Bedminster, NJ 07921 MARK C, FLEMING

(908) 532-1850 WILMER CUTLER PICKERING

HALE AND DORR LLP

1875 Pennsylvania Ave., N.W.

Washington, DC 20006

(202) 663-6000

JANUARY 2007

(fh A.& <a 54 ee

BS a tie

ieee Sue" ; I

es al EE Oe eae

‘J ae as vs igh cS at pn

tan: * oo

ae

+s

Ks .

;

7.)

J

}

'

a.

;

i

; 7

is

wa 4

J “iL

Py <P F |

> -.

= et)

i

a A =) ae

- = ona

F Ss ; i=

Me fry ie > oo «

aif a) ore, *

}

~ -} - a]