Amicus Curiae Brief — Mordechai Korf, et al., Petitioners v. United States
Supreme Court briefMay 20, 2022
Ask Donna
What actually matters in this document.
Text
No. 21-1364
In the
Supreme Court of the United States
MORDECHAI KORF ET AL.,
Petitioners,
v.
UNITED STATES OF AMERICA,
Respondent.
ON PETITION FOR A WRIT OF CERTIORARI TO THE
UNITED STATES COURT OF APPEALS FOR THE
ELEVENTH CIRCUIT
Brief of Amici Curiae Professors of Criminal
Law and Legal Ethics in Support of Petitioners
JONATHAN I. KRAVIS
Counsel of Record
BRENDAN B. GANTS
MUNGER, TOLLES & OLSON LLP
601 Massachusetts Ave. NW
Suite 500E
Washington, D.C. 20001
(202) 220-1100
Jonathan.Kravis@mto.com
Counsel for Amici Curiae
i
TABLE OF CONTENTS
Page
INTEREST OF AMICI CURIAE ................................. 1
SUMMARY OF ARGUMENT ..................................... 3
ARGUMENT ................................................................ 4
I.
The Use of Government “Filter Teams”
Imperils Client Confidences and
Weakens the Actual and Perceived
Fairness of Judicial Proceedings. ..................... 4
II.
Filter Teams in Numerous Cases Have
Failed to Protect Defendants’ Rights in
Privileged Material. ........................................ 10
III.
Alternative Approaches That Would
Avoid These Problems are Available and
Should Be Adopted. ......................................... 21
CONCLUSION .......................................................... 24
ii
TABLE OF AUTHORITIES
Page
FEDERAL CASES
Brady v. Maryland,
373 U.S. 83 (1963) .......................................... 10, 13
In re Grand Jury Proc.,
220 F.3d 568 (7th Cir. 2000) .................................. 7
In re Grand Jury Subpoenas,
454 F.3d 511 (6th Cir. 2006) .................... 6, 7, 8, 23
Harbor Healthcare Sys., L.P. v. United
States,
5 F.4th 593 (5th Cir. 2021) ......................... 5, 11, 12
Hickman v. Taylor,
329 U.S. 495 (1947) ................................................ 5
In re Kellogg Brown & Root, Inc.,
756 F.3d 754 (D.C. Cir. 2014) ................................ 8
Klitzman, Klitzman & Gallagher v.
Krut,
744 F.2d 955 (3d Cir. 1984) .................................. 22
In re Search Warrant for Law Offices
Executed on March 19, 1992,
153 F.R.D. 55 (S.D.N.Y. 1994) ........................... 8, 9
In re Search Warrant Issued June 13,
2019,
942 F.3d 159 (4th Cir. 2019) ........................ 5, 7, 22
iii
TABLE OF AUTHORITIES
Page
In re Search Warrants Executed on April
9, 2018,
No. 18-MJ-3161 (S.D.N.Y.)................................... 22
SEC v. Lek Securities Corp.,
No. 17-cv-1789, 2018 WL 417596
(S.D.N.Y. Jan. 16, 2018) ................................. 16, 17
Swidler & Berlin v. United States,
524 U.S. 399 (1998) .............................................. 23
United States v. Avenatti,
No. 8:19-cr-061 (C.D. Cal. Aug. 28,
2021)...................................................................... 13
United States v. Black,
No. 16-cr-20032, 2017 WL 2151861
(D. Kan. May 17, 2017) ........................................ 17
United States v. Carter,
429 F. Supp. 3d 788 (D. Kan. 2019),
order vacated in part on other
grounds, No. 16-cr-20032-02, 2020
WL 430739 (D. Kan. Jan. 28, 2020) ............... 17, 18
United States v. DeLuca,
663 F. App’x 875 (11th Cir. 2016) ........................ 18
United States v. Elbaz,
396 F. Supp. 3d 583 (D. Md. 2019) ................ 13, 14
United States v. Esformes,
No. 16-cr-20549, 2018 WL 5919517
(S.D. Fla. Nov. 13, 2018) .......................... 14, 15, 16
iv
TABLE OF AUTHORITIES
Page
United States v. Gallego,
No. 4:18-cr-1537, 2018 WL 4257967
(D. Ariz. Sept. 6, 2018) ..................................... 9, 22
United States v. Kaplan,
No. 02-cr-883, 2003 WL 22880914
(S.D.N.Y. Dec. 5, 2003) ......................................... 20
United States v. Nobles,
422 U.S. 225 (1975) ................................................ 5
United States v. Noriega,
764 F. Supp. 1480 (S.D. Fla. 1991) ................ 20, 21
United States v. Pedersen,
No. 12-cr-431, 2014 WL 3871197 (D.
Or. Aug. 6, 2014)............................................. 19, 20
United States v. Rayburn House Office
Building, Room 2113,
497 F.3d 654 (D.C. Cir. 2007) ................................ 6
United States v. SDI Future Health, Inc.,
464 F. Supp. 2d 1027 (D. Nev. 2006) ..................... 7
United States v. Stewart,
No. 02-cr-3, 2002 WL 1300059
(S.D.N.Y. June 11, 2002) ...................................... 22
United States v. Sullivan,
No. 17-cr-104, 2020 WL 1815220 (D.
Haw. Apr. 9, 2020) .......................................... 12, 13
v
TABLE OF AUTHORITIES
Page
United States v. Taylor,
764 F. Supp. 2d 230 (D. Me. 2011) ......................... 7
United States v. Zolin,
491 U.S. 554 (1989) ........................................ 22, 23
Upjohn Co. v. United States,
449 U.S. 383 (1981) ...................................... 5, 6, 23
STATE CASES
O’Connor v. Johnson,
287 N.W.2d 400 (Minn. 1979) ................................ 9
Preventive Med. Assocs., Inc. v.
Massachusetts,
992 N.E.2d 257 (Mass. 2013) ................................. 9
FEDERAL STATUTES
42 U.S.C. § 2000aa-11(a) ............................................. 6
OTHER AUTHORITIES
Roland Behm et. al., “Trust Us:” Taint
Teams and the Government’s Peek at
Your Company’s Privileged
Documents, ACC Docket, June 2010 ................... 10
vi
TABLE OF AUTHORITIES
Page
Ellen S. Podgor & Wilma F. Metcalf, The
Fox Guarding the Henhouse:
Government Review of AttorneyClient Privileged Material in White
Collar Cases, 103 B.U. L. Rev.
(forthcoming 2023), available at
https://papers.ssrn.com/sol3/
papers.cfm?abstract_id=4087032 .......................... 1
Paul R. Rice, Attorney-Client Privilege in
the U.S. § 11:19 (2d ed. 2021) ................................ 8
U.S. Dep’t of Just., Just. Manual § 913.420 (2021), available at
https://www.justice.gov/jm/jm-913000-obtaining-evidence#9-13.420 ...................... 6
1
INTEREST OF AMICI CURIAE 1
Ellen S. Podgor is the Gary R. Trombley Family
White-Collar Crime Research Professor of Law at
Stetson University College of Law. A former deputy
prosecutor and former criminal defense attorney, Professor Podgor teaches and writes in the areas of
white-collar crime, criminal law, and criminal procedure. She has authored or co-authored numerous
books and more than 70 law review articles and essays, including a forthcoming article about filter
teams and alternative approaches to reviewing attorney-client privileged material. 2
Gregory M. Gilchrist is Professor of Law at the
University of Toledo, where he teaches and writes in
the areas of criminal law, criminal procedure, and
white-collar crime. Professor Gilchrist has been a
criminal defense lawyer in private practice and in the
Office of the Federal Public Defender, and continues
to represent individuals in federal criminal cases today.
David Kwok is George A. Butler Research Professor and Associate Professor at the University of Houston Law Center, where he is also co-director of the
Pursuant to Supreme Court Rule 37.6, counsel for amici certify
that no counsel for any party authored this brief in whole or in
part, and no party or counsel for a party, or any other person
other than amici or their counsel, made a monetary contribution
intended to fund the preparation or submission of this brief.
Counsel for all parties received timely notice of amici’s intent to
file this brief, and consented in writing to its filing.
1
See Ellen S. Podgor & Wilma F. Metcalf, The Fox Guarding the
Henhouse: Government Review of Attorney-Client Privileged Material in White Collar Cases, 103 B.U. L. Rev. (forthcoming
2023), available at https://papers.ssrn.com/sol3/papers.cfm?abs
tract_id=4087032.
2
2
Criminal Justice Institute. Professor Kwok’s teaching and scholarship is focused on criminal law, criminal procedure, and white-collar crime.
Cliff Sloan is Distinguished Visitor from Practice
at Georgetown University Law Center, where he
teaches classes in criminal justice and constitutional
law. He has represented both the government and
defendants in criminal cases.
Ellen Yaroshefsky is the Howard Lichtenstein
Distinguished Professor in Legal Ethics and Professor of Law at the Maurice A. Deane School of Law at
Hofstra University. Professor Yaroshefsky teaches
and writes about legal ethics issues, including in particular those that arise in the context of criminal law.
She also counsels lawyers, law firms, and bar associations regarding legal ethics issues.
3
SUMMARY OF ARGUMENT
The use of government “filter teams” to review
assertedly privileged material seized by the
government raises grave concerns about the
protection of client confidences. As numerous courts
have recognized, any such process fundamentally
undermines the attorney-client privilege and the
work product doctrine.
Moreover, the practical
reality is that government filter teams are inclined to
take a relatively narrow view of privilege, suffer from
a lack of information needed to reliably assess
privilege, and are prone to errors—which, when they
happen, can be difficult or impossible for anyone
outside the government to detect. Such processes do
not adequately protect the attorney-client privilege,
and they undermine public confidence in the
fundamental fairness of the judicial process. These
concerns are particularly acute where, as here, the
search includes the files and electronic devices of
attorneys
representing
the
subject
of
the
investigation.
These serious problems are not merely theoretical,
as numerous examples of filter-team failures show.
In cases across the nation, government filter teams
have failed to screen documents for privilege or failed
to identify privileged material they did screen, with
the result that prosecution teams have received,
reviewed, and in some cases used privileged material
in their investigations. Even worse, in many of these
cases the government failed to recognize the problem
or failed to report it to the court or to defense
counsel—which raises the realistic prospect that the
failures of which the public is aware may be merely
the tip of the iceberg.
4
Recognizing these inherent flaws in filter teams,
courts increasingly have adopted alternative
approaches that better protect client confidences.
The Eleventh Circuit’s decision in this case
erroneously rejected those approaches and failed to
recognize the serious problems with filter teams.
This Court should grant certiorari to provide crucial
guidance to courts nationwide on this important
issue.
ARGUMENT
I.
The Use of Government “Filter Teams”
Imperils Client Confidences and Weakens
the Actual and Perceived Fairness of
Judicial Proceedings.
As the Petition explains, the Courts of Appeals are
divided over how to implement this Court’s existing
precedent regarding the attorney-client privilege and
work product doctrine in the context of searches and
seizures, and this Court’s review is needed to resolve
that disagreement. Numerous courts have recognized, correctly, that permitting government filter
teams to review assertedly confidential attorneyclient material undermines those essential protections, creates a significant risk that privileged documents will be disclosed to the investigation team, and
jeopardizes public confidence in the fairness of judicial proceedings.
1. Any review of a criminal defendant’s privileged
material by government agents as part of a filter
team undermines the fundamental purpose of the attorney-client privilege—“to encourage full and frank
communication between attorneys and their clients
and thereby promote broader public interests in the
observance of law and administration of justice.”
5
Upjohn Co. v. United States, 449 U.S. 383, 389 (1981).
“The whole point of privilege is privacy.” Harbor
Healthcare Sys., L.P. v. United States, 5 F.4th 593,
599 (5th Cir. 2021).
The same is true of the work product doctrine,
which “most frequently is asserted as a bar to discovery in civil litigation,” but which plays an “even more
vital” role “in assuring the proper functioning of the
criminal justice system.” United States v. Nobles, 422
U.S. 225, 237-38 (1975). As this Court has long emphasized, if attorney work product were discoverable
by the client’s adversary, “much of what is now put
down in writing would remain unwritten,” and “the
interests of the clients and the cause of justice would
be poorly served.” Hickman v. Taylor, 329 U.S. 495,
511 (1947).
Together, these principles cordon off a “privileged
area” in which lawyer and client may plan and present a defense. In re Search Warrant Issued June 13,
2019, 942 F.3d 159, 174 (4th Cir. 2019) (citation omitted). As courts have increasingly come to recognize,
the effect of using a government filter team is to invade that privileged area. See, e.g., id. at 174-75.
Clients will be less likely to share confidences with
their attorneys, and attorneys will be less likely to
create work product, if they know that government
agents may review their materials. That the government agents reviewing privileged communications
may be segregated from one particular investigation
is cold comfort—especially because the government
reserves the right, consistent with the plain-view doctrine, to use in other criminal inquiries any information that its agents learn as part of a filter team.
See id. at 183 & n.21.
Filter-team processes therefore are likely to have a
chilling effect on attorney-client communications and
6
on attorney-led factual investigations, including internal investigations into alleged corporate wrongdoing. Such investigations generate privileged materials that would be especially useful to government investigators, if they were allowed access to them. See
Upjohn, 449 U.S. at 396. Notably, in one criminal
case that involved a raid on the defendant’s congressional office, the D.C. Circuit Court of Appeals rejected use of a filter team to review documents potentially protected by the Speech or Debate Clause out of
concern that the process could “chill the exchange of
views with respect to legislative activity.” United
States v. Rayburn House Office Building, Room 2113,
497 F.3d 654, 661 (D.C. Cir. 2007). So too, the use of
filter teams to review potentially privileged documents threatens to chill the exchange of information
and views between attorney and client.
The availability of filter-team processes only in the
context of a search and seizure may also provide a
perverse incentive to proceed in that manner. Respect for the principles underlying the attorney-client
privilege dictates that, where prosecutors must obtain evidence from lawyers or represented parties,
they should use the least intrusive practicable
means—a proposition that in some instances is codified by statute and in the Department of Justice
Manual. See 42 U.S.C. § 2000aa-11(a); U.S. Dep’t of
Just., Just. Manual § 9-13.420 (2021), available at
https://www.justice.gov/jm/jm-9-13000-obtainingevidence#9-13.420. But when the government does
proceed under a means less intrusive than search and
seizure, such as by subpoena, courts have been unwilling to approve filter-team processes. See In re
Grand Jury Subpoenas, 454 F.3d 511, 523 (6th Cir.
2006) (rejecting such a process because “the government does not actually possess the potentially-
7
privileged materials here, so the exigency typically
underlying the use of taint teams is not present”);
United States v. Taylor, 764 F. Supp. 2d 230, 234 (D.
Me. 2011) (noting that courts have generally required
preliminary review by defense counsel “where the
government has not yet obtained the records”); United States v. SDI Future Health, Inc., 464 F. Supp. 2d
1027, 1037 (D. Nev. 2006) (similar). Insofar as the
government views filter-team procedures as advantageous—which it evidently does in many cases—this
incentive structure is troubling.
2. In addition, the use of government filter-team
agents to screen potentially privileged material is beset with numerous practical problems.
For one thing, “[i]t is reasonable to presume that
the government’s taint team might have a more restrictive view of privilege than [the defendant’s] attorneys. In re Grand Jury Subpoenas, 454 F.3d at
523. That more restrictive view, understandable
“given their prosecutorial interests in pursuing the
underlying investigations,” may “cause privileged
documents to be misclassified and erroneously provided to an investigation or prosecution team.” In re
Search Warrant Issued June 13, 2019, 942 F.3d at
177.
This problem is exacerbated by the fact that the filter team frequently lacks information necessary to
assess privilege. “The inquiry into whether documents are subject to a privilege is a highly factspecific one” that requires taking context into account. In re Grand Jury Proc., 220 F.3d 568, 571 (7th
Cir. 2000). For example, a document not sent to or
from a lawyer could be privileged, depending on the
nature of the document, the purpose(s) for which it
was created, and other factual circumstances, information that a filter team is unlikely to possess. See,
8
e.g., In re Kellogg Brown & Root, Inc., 756 F.3d 754,
760 (D.C. Cir. 2014). As one treatise explains:
The privileged status of many communications is not apparent on their face.
This, of course, is why privilege proponents generally have to supply supporting materials to opposing parties and
the reviewing judicial officers when their
claims are made. None of this often essential information can even be made
available to the members of the taint
team because they cannot reveal the
contents of any of the seized communications to those who could supply it. As a
consequence, this taint team process will
necessarily result in a large number of
false negatives—privileged documents
that are not recognized as such.
Paul R. Rice, Attorney-Client Privilege in the U.S.
§ 11:19 (2d ed. 2021). A related problem, which occurred in one case discussed below, arises when filterteam members also make determinations about
which documents are relevant or exculpatory.
For these and other reasons, when government filter teams review assertedly privileged material, mistakes—and sometimes misconduct—are too often the
result. Filter teams in numerous cases “have been
implicated in the past in leaks of confidential information to prosecutors. . . . [H]uman nature being
what it is, occasionally some taint-team attorneys
will make mistakes or violate their ethical obligations.” In re Grand Jury Subpoenas, 454 F.3d at 523;
see infra Part II. Reliance on such procedures, “especially in the context of a criminal prosecution, is highly questionable, and should be discouraged,” as leaks
may indeed be “inevitable.” In re Search Warrant for
9
Law Offices Executed on March 19, 1992, 153 F.R.D.
55, 59 (S.D.N.Y. 1994).
The practical problems with filter teams are not
limited to the investigation at issue. Members of the
filter team may become involved in another investigation or prosecution in which information they
learned on the filter team becomes relevant, whether
or not they realize it. In light of its fact-specific nature, this issue cannot be addressed through government screening procedures, and the defendant may
never learn sufficient information about the two investigations to detect or raise the issue. Even the
former filter-team member might not recall when or
how she learned the information. But once information protected by the attorney-client privilege or
work product doctrine “is revealed to the police, the
privileges are lost, and the information cannot be
erased from the minds of the police.” O’Connor v.
Johnson, 287 N.W.2d 400, 405 (Minn. 1979).
3. Finally, as courts have noted repeatedly, the
use of filter teams undermines public confidence in
the fundamental fairness of judicial proceedings. As
one court observed, “The appearance of Justice must
be served, as well as the interests of Justice. It is a
great leap of faith to expect that members of the general public would believe any such [ethical] wall
would be impenetrable; this notwithstanding our own
trust in the honor of an AUSA.” In re Search Warrant for Law Offices Executed on March 19, 1992, 153
F.R.D. at 59; see also United States v. Gallego, No.
4:18-cr-1537, 2018 WL 4257967, at *2 (D. Ariz. Sept.
6, 2018) ([E]ven if no leaks occur, the use of walled-off
taint teams undermines the appearance of fairness
and justice.”); Preventive Med. Assocs., Inc. v. Massachusetts, 992 N.E.2d 257, 272 (Mass. 2013) (“[U]se of
an independent special master offers a far greater
10
appearance of impartiality and protection against
unwarranted disclosure and use of an indicted defendant’s privileged communications.”).
This problem is significantly compounded by the
fact that the filter-team process is conducted behind
closed doors under the government’s own supervision,
making it difficult if not impossible for anyone outside the government to know whether an unauthorized disclosure of privileged information has occurred.
“[T]he secretive nature of cases in which taint teams
are often employed leave many accounts of botched
procedures that result in privilege leaks relegated to
rumor or sealed judicial memoranda.” Roland Behm
et. al., “Trust Us:” Taint Teams and the Government’s
Peek at Your Company’s Privileged Documents, ACC
Docket, June 2010, at 74, 82-83. In short, the use of
filter teams calls into question the fairness of judicial
proceedings where privilege is at issue.
II.
Filter Teams in Numerous Cases Have
Failed to Protect Defendants’ Rights in
Privileged Material.
Despite the lack of transparency that makes filterteam failures so difficult to identify, some severe failures to protect defendants’ privileged material have
come to light over the years, along with at least one
recent case in which a filter team failed to protect the
defendant’s rights under Brady v. Maryland, 373 U.S.
83 (1963). Disturbingly, these failures typically were
revealed by accident rather than through any forthright admission by the government—and in some
cases, the government refused to acknowledge any
error even as it faced sanctions for misconduct.
Examples of the failures of filter teams include:
11
1.
Harbor Healthcare System v. United States:
After Harbor Healthcare System (“Harbor”) became
the subject of two qui tam lawsuits alleging False
Claims Act violations, the company received investigative demands from the U.S. Department of Health
and Human Services and the Department of Justice
(“DOJ”) Civil Division. The company responded to
the demands, led by its compliance director and outside counsel. See Harbor Healthcare Sys., L.P. v.
United States, 5 F.4th 593, 595 (5th Cir. 2021). The
local U.S. Attorney’s Office (“USAO”) subsequently
executed a search warrant on Harbor offices, seizing
huge swaths of data, including recent privileged
communications with outside counsel regarding the
investigative demands. See id. at 596.
As in the case now before this Court, in Harbor the
government assembled a filter team from a different
division of the USAO handling the investigation, received from the company a list of lawyers and law
firms, and set up a process by which the company
should be able to challenge the filter team’s privilege
determinations. Yet when the company did so, the
government refused to respond, and failed to return
or delete privileged material. Eventually, Harbor received a list of documents the government had already transferred to civil and criminal investigators,
which included “a significant number of privileged
documents.” Id. at 597.
The Fifth Circuit found that the government
“show[ed] a ‘callous disregard’ for Harbor’s rights”
and “made no attempt to respect Harbor’s right to attorney-client privilege in the initial search.” Id. at
599. Moreover, “by its treatment of Harbor’s privileged materials after the search,” the government
“further disregarded Harbor’s rights” when it refused
12
to destroy or return copies of documents that it
agreed were privileged. Id.
2.
United States v. Sullivan:
Because defendant Leihinahina Sullivan was represented by counsel when the government executed a
search warrant at her residence, the government established a filter team to review all seized materials.
That team failed to identify as privileged, and therefore sent to the prosecution team, four photographs of
defense “strategy boards” related to the pending case.
United States v. Sullivan, No. 17-cr-104, 2020 WL
1815220, at *6 (D. Haw. Apr. 9, 2020). The court
found that a filter-team agent was unable to open the
relevant computer files and so, rather than taking
even “minimal” extra steps to protect the defendant’s
rights, simply “presumed the documents were not
privileged and thus provided them to the prosecution
team.” Id. at *8-9. The defendant, who represented
herself pro se, identified the privileged documents
while reviewing discovery in pretrial detention and
brought them to the court’s attention. Id. at *4 &
n.11.
The court observed that the filter-team agent
“demonstrat[ed] a total disinterest in both the rights
of Defendant and the court’s expectation that the
taint team would fulfill its obligation to the court,”
even after learning that he had violated both. Id. at
*8. The government’s filings similarly reflected an
“astonishing” and “disappointing lack of recognition
of this wrongdoing.” Id. at *9. Indeed, the government, while recognizing that the filter team had disclosed privileged documents to the prosecution team,
argued that “there was no improper disclosure” because technically “there was no failure to follow taint
review protocols,” an argument the court rightly
characterized as “shocking.” Id. at *9 & n.20.
13
The court sanctioned the government for the “reckless and grossly negligent conduct” of the filter team
and found that “the United States’ conduct cannot be
described as a mistake or honest disagreement of
opinion; instead, it demonstrated a clear lack of concern for Defendant’s rights and its obligations to this
court.” Id. at *10.
3.
United States v. Avenatti:
Invasion of the attorney-client privilege is not the
only harm that can result from the use of a filter
team, as the recent mistrial in the Avenatti case illustrates.
In declaring a mistrial, the district court found that
the filter team had failed to provide to the prosecution team—and consequently, the government failed
to provide to Avenatti—material the government was
required to disclose under Brady. The court found
that “the government was fully on notice of the significance of” the material, but that through “inadvertence and a failure to appreciate what was there,” it
failed to make the required disclosure, thus depriving
Avenatti of information he could have used in his
opening statement and in cross-examining government witnesses. Trial Transcript at 57, 62-63, United
States v. Avenatti, No. 8:19-cr-061 (C.D. Cal. Aug. 28,
2021), ECF No. 792.
Notably, although in Avenatti the defendant knew
that the relevant documents were missing, in many
cases defendants may be unaware of the existence of
particular Brady material that falls through the
cracks of the government’s faulty filter-team process.
4.
United States v. Elbaz:
In Elbaz, the government, struggling to meet discovery deadlines, took a series of shortcuts that led to
14
privileged material being disclosed by the filter team
to the prosecution team. First, the prosecution team
requested the full original contents of one seized hard
drive before the filter team had completed its review—and then failed to notice the filter team’s
warning that it could contain privileged materials.
See United States v. Elbaz, 396 F. Supp. 3d 583, 58889 (D. Md. 2019). Additionally, the prosecution team
inferred, incorrectly, that the contents of one seized
device would not contain privileged material, and so
did not run it through the filter process. See id. at
589. Upon further review, the government discovered
that the prosecution team also had access to
WhatsApp chats and Hebrew-language documents
that had not been reviewed by the filter team. See id.
at 589-90. Cumulatively, these errors gave the prosecution team access to thousands of potentially privileged documents for approximately five months, more
than 100 of which members of that team viewed or
were presumed to have viewed. See id. at 590.
5.
United States v. Esformes:
This health care fraud case involved a remarkable
variety of filter-team problems, all of which the district court found had occurred despite the government acting in good faith.
First, although the government’s filter-team protocol required it to use “non-case agents” to execute an
office search, the government “relied on several
agents for the search who had participated in other
health care fraud cases that bear some relationship to
the Esformes case or who were later used in the underlying Esformes investigation.” United States v.
Esformes, No. 16-cr-20549, 2018 WL 5919517, at *23
(S.D. Fla. Nov. 13, 2018). The district court noted
that the use of these agents “raises questions about
15
their independence and effectiveness as members of
the filter team.” Id.
The government’s selection of these agents was only
the beginning of its noncompliance with the filterteam protocol, as “agents were either provided inadequate instructions or ignored those instructions”
throughout a search that was “clumsy” and “borderline incompetent.” Id. “Hundreds of documents,
clearly prepared by law firms and/or marked ‘privileged and confidential’ or ‘attorney/client privilege’ or
‘work product privileged’ or ‘legal’ were not segregated and placed in the ‘taint box’ and were, thus, within
the boxes provided to the prosecution team.” Id. at
21. And like the government’s “preparation for and
the methods used during” the search, those it used
“after the search were sloppy and ineffective.” Id. at
23.
Consequently, the prosecution team not only reviewed privileged documents but also used them in a
reverse proffer with one witness and debriefings of
another. See id. The prosecution team continued to
review and use these documents even after being
alerted to the privilege issue by one of the witnesses—and when that team later came across a document that was potentially privileged on its face, it did
not inform the court or the defense. See id. at 24.
These issues came to light only after the defense
reviewed the seized materials, identified privileged
documents, and alerted the government—at which
point the prosecution team represented inaccurately
to defense counsel that the seized materials were “being reviewed by a filter team and not by the prosecution team.” Id.
The district court declined to adopt the magistrate
judge’s finding that the prosecutors acted in bad
16
faith, determining instead that “their execution of
their duties was often sloppy, careless, clumsy, ineffective, and clouded by their stubborn refusal to be
sufficiently sensitive to issues impacting the attorney
client privilege.” Id. at 34. In particular, the prosecution team’s “myopic view of [the defendant’s lawyer] as a criminal and not an attorney skewed their
reaction to, and blurred their ability to see, the potential for privilege.” Id. at 32.
The district court suppressed some evidence but
denied other remedies, finding that the defense failed
to meet its burden to establish prejudice. The court
acknowledged that the government “attempted to use
the information contained in [the defendant’s privileged] documents to gather more information from”
the defendant’s confidant and his assistant, but nevertheless the court was “unconvinced” that the government gained a “real advantage” from their use.
Id. at 31.
6.
SEC v. Lek Securities Corporation:
To support the SEC’s investigation of a securities
company for manipulative trading, the New Jersey
USAO established a filter-team process to cull potentially privileged documents using a defined set of
privilege terms. Yet supposedly “filtered” materials
provided to the SEC repeatedly included documents
that hit on the privilege terms, and an SEC attorney
determined that dozens of those documents were potentially privileged. See SEC v. Lek Securities Corp.,
No.17-cv-1789, 2018 WL 417596, at *1 (S.D.N.Y. Jan.
16, 2018). Moreover, when the SEC asked the USAO
filter team for documents in native format, the filter
team failed to screen the attachments to documents
that hit on privilege terms. See id. at *2. These failures to implement a relatively straightforward filterteam process resulted in the SEC investigative team
17
receiving a 15-page draft timeline of events that the
company’s owner sent to his attorney (which the SEC
acknowledged reviewing), as well as documents prepared by the attorney or his paralegal (which were
clearly identified in a footer as privileged attorney
work product) describing the “key” individuals and
entities in the case. See id. at *3-4.
7.
United States v. Black:
During an investigation into alleged drug smuggling in a federal detention facility, the government
obtained audio recordings of prisoners’ phone calls to
their attorneys. The government created a filter
team to review other materials but did not ask that
team to review the audio recordings. Instead, the
lead prosecuting attorney distributed the recordings
“indiscriminately,” even after learning that investigators had inadvertently listened to the beginning of at
least one attorney-client call. United States v. Black,
No. 16-cr-20032, 2017 WL 2151861, at *7 (D. Kan.
May 17, 2017). The district court determined that
she did so because—despite a contrary opinion from
DOJ’s Professional Responsibility Advisory Office—
the government “unilaterally decided the telephone
calls were not privileged, and did so without notice to
the Court or the parties.” Id. at *8.
Following an investigation by a special master, the
court found that at least two prosecutors “had knowingly and intentionally listened to attorney-client
phone calls in one or more of their cases.” United
States v. Carter, 429 F. Supp. 3d 788, 824 (D. Kan.
2019), order vacated in part on other grounds, No. 16cr-20032-02, 2020 WL 430739 (D. Kan. Jan. 28,
2020). The extent and scope of the government’s intrusion on the privilege remained unclear because the
government defied the court’s orders to preserve and
produce relevant documents and generally refused to
18
cooperate with the investigation. See id. at 866-67.
Ultimately, the court found that the government had
committed “systemic prosecutorial misconduct” and
then “responded by minimizing the seriousness of its
conduct and delayed and obfuscated th[e] investigation” of that conduct. Id. at 903. Litigation over that
misconduct continues today.
8.
United States v. DeLuca:
After the FBI searched the offices of defendant Stephen DeLuca’s company, the parties stipulated to a
process under which a filter team would review
DeLuca’s communications to or from his attorneys
and give DeLuca any documents deemed not privileged so that he could decide whether to challenge
that determination before a magistrate. A member of
the filter team, however, “unilaterally decided that
the stipulation was not in effect” and so gave the
prosecution team access to communications that he
deemed not privileged, without notice to the defense
or the court. United States v. DeLuca, 663 F. App’x
875, 877 (11th Cir. 2016). The defense discovered
this violation only when, on the eve of trial, the government filed an amended exhibit list that included
one of DeLuca’s emails to his attorney. See id.
The district court found that the government had
violated the attorney-client privilege by viewing that
email and other attorney-client communications, but
denied relief on the ground that DeLuca failed to
meet his burden to show prejudice. See id. at 878. As
the Eleventh Circuit noted in affirming, “showing
how the government used certain information within
its control as part of a criminal investigation has always been an uphill battle.” Id. at 881.
19
9.
United States v. Pedersen:
In this capital case, government violations of the
attorney-client privilege, resulting from both intentional conduct and neglect, continued for years because the filter-team protocol was both deficient and
not followed, “because nobody notified defense counsel regarding these problems, and because nobody
raised these issues with the court despite many opportunities to do so.” United States v. Pedersen, No.
12-cr-431, 2014 WL 3871197, at *26 (D. Or. Aug. 6,
2014).
The government’s numerous violations of the filterteam protocol included failing to provide the filter
team with intercepted legal mail despite explicit instructions to do so, see id. at *31, and failing to run
through the filter-team process intercepted audio of
attorney-client calls, some of which included “significant substantive content regarding a number of issues including the facts of the case, legal issues pertaining to the case, [and] defense strategy,” id. at *21.
The government also failed to maintain a clear separation between the filter team and the prosecution
team—one AUSA’s dual role was “anathema to the
very purpose of a taint team.” Id. at *17, 30.
The court noted that the “most troubling aspect” of
the government’s conduct was that, despite being
aware of the problems, the government “did not alert
the court . . . on its own.” Id. at *32. Indeed, the
court observed, “It is unclear when, if ever, the government would have raised these issues on its own.
Rather, it was the accidental provision of privileged
calls through discovery and the very hard work of defense counsel that ultimately resulted in the exposure
of the conduct in this case.” Id. Partly for that reason, the court expressed concern that the govern-
20
ment’s transgressions in the case were “likely to recur
absent corrective action.” Id. at *2.
10. United States v. Kaplan:
The government, suspecting that defendant Solomon Kaplan was perpetrating fraud through his law
practice, searched Kaplan’s office and seized numerous files. The affidavit in support of the search warrant detailed procedures through which an “ethical
wall” attorney would screen seized documents for
privilege. United States v. Kaplan, No. 02-cr-883,
2003 WL 22880914, at *1 (S.D.N.Y. Dec. 5, 2003).
The court found it “apparent that the procedures
used by the Government in this case were of little use
in protecting any privileged materials seized by the
Government.” Id. at *11. Contrary to the procedures
the government affirmed to the magistrate judge
would be used, an FBI case agent—not a member of
the “ethical Wall Team”—was given access to privileged materials and allowed to make his own determination whether the crime-fraud exception applied.
Id. The court noted that the breakdown of the filterteam procedures in this case “raises serious concerns
about the admissibility of information gained in the
investigation of leads developed as a result of review
of materials that ultimately are determined to be
privileged, and eviscerates any claim that an ‘ethical
wall team’ within the Government effectively screens
the prosecution team from privileged materials.” Id.
11. United States v. Noriega:
Perhaps the most famous example of filter-team
failure occurred in the government’s prosecution of
former Panamanian dictator Manuel Noriega. After
the government obtained audio recordings of Noriega’s prison phone calls, it developed procedures
pursuant to which the recordings would be screened
21
for privilege by a DEA agent not involved in the prosecution. See United States v. Noriega, 764 F. Supp.
1480, 1483 (S.D. Fla. 1991). This filter agent failed to
screen many of the recordings, however, and agents
assisting the prosecution received and reviewed numerous attorney-client conversations, including a
conversation about two potential government witnesses that a government witness himself reviewed.
After that witness conveyed the substance of the privileged conversation in a memo he prepared for the
prosecution’s use, the lead prosecutor asked the filter
agent to check whether the relevant audiotape contained any attorney-client conversations—the filter
agent erroneously reported back that it did not. See
id. at 1483-84.
The government’s review of Noriega’s privileged
conversations with his attorney came to light only
when CNN somehow obtained copies of the recordings. The district court observed that it was “unclear
why” Noriega’s privileged conversations with his defense team “were not recognized as such by the outside screeners and DEA agents reviewing the tapes,
or by the agents briefed by” the government’s witness. Id. at 1489.
III.
Alternative Approaches That Would
Avoid These Problems are Available and
Should Be Adopted.
Alternative approaches for reviewing assertedly
privileged material after a search and seizure are
available, have been adopted in numerous cases, and
are more protective of client confidences than filterteam procedures.
The most common alternative approach to a government filter team is for the district court to appoint
22
a special master or magistrate judge to make privilege determinations. Importantly, under this approach—unlike that adopted below—the government
has no opportunity to review the content of documents that the adjudicator later determines are privileged. Consequently, there is no risk that filter-team
agents will—inadvertently or otherwise—disclose
privileged documents to the prosecution team, or
share information they have learned in those documents, or use that information in another investigation. This approach, with its neutral arbiter and other attendant safeguards, is more protective of privilege and more likely to engender (and be worthy of)
public confidence in the administration of justice.
That alternative is the approach adopted, for example, in the recent cases involving Michael Cohen
and Rudy Giuliani. See Order of Appointment, In re
Search Warrants Executed on April 9, 2018, No. 18MJ-3161 (S.D.N.Y. April 27, 2018), ECF No. 30;
Transcript of Hearing at 88, In re Search Warrants
Executed on April 9, 2018, No. 18-MJ-3161 (S.D.N.Y.
Sept. 18, 2018), ECF No. 104. Circuit courts have
recommended this approach, and district courts have
adopted it, for decades in other cases. See In re
Search Warrant Issued June 13, 2019, 942 F.3d at
181; Klitzman, Klitzman & Gallagher v. Krut, 744
F.2d 955, 962 (3d Cir. 1984); Gallego, 2018 WL
4257967, at *3; United States v. Stewart, No. 02-cr-3,
2002 WL 1300059, at *10 (S.D.N.Y. June 11, 2002).
The particular approach proposed below by Petitioners—preparation of a privilege log by the defendant,
followed by judicial resolution of any disputes that
may arise—is common to civil litigation, and in the
criminal context would simply require the government to make the initial showing required by United
23
States v. Zolin, 491 U.S. 554 (1989), before the district court could review documents in camera.
As the various cases adopting alternatives to filterteam procedures illustrate, these alternatives are
well-established and eminently workable. Indeed,
the government itself has proposed engaging a special master in some cases, including the Giuliani
case. And of course, the government regularly invokes the process prescribed by Zolin in civil cases,
where more than three decades of experience have
amply demonstrated its viability.
Concerns about administrability and cost provide
no sound justification for adopting the repeatedly
failed filter-team procedures. Filter-team alternatives are not only widely used but also easily adaptable. Insofar as timeliness is a concern, for example, a
special master’s review can be conducted, and productions made, on a rolling basis. See, e.g., In re
Grand Jury Subpoenas, 454 F.3d at 524. Moreover,
costs—for which the government will ultimately be
responsible regardless of whether government agents
or a special master reviews the documents—can be
minimized by allowing for some assistance from the
parties, without permitting government agents to review the content of privileged documents. Moreover,
this Court has emphasized that “considerations of
convenience do not overcome the policies served by
the attorney-client privilege,” Upjohn, 449 U.S. at
396, and has repeatedly declined to justify intrusions
on the privilege based on prosecutorial interests such
as these. See, e.g., id.; Swidler & Berlin v. United
States, 524 U.S. 399, 410 (1998).
*
*
*
The use of filter teams fundamentally and unnecessarily intrudes on criminal defendants’ attorney-
24
client privilege and work product doctrine, and has
led to severe violations of those protections in numerous cases. The time has come for this Court to take a
close look at those procedures and provide needed
guidance.
CONCLUSION
For the foregoing reasons, amici respectfully
submit that the Court should grant the petition for a
writ of certiorari.
May 20, 2022
Respectfully submitted,
COUNSEL FOR AMICI CURIAE
JONATHAN I. KRAVIS
Counsel of Record
BRENDAN B. GANTS
MUNGER, TOLLES & OLSON LLP
601 Massachusetts Ave. NW
Suite 500E
Washington, D.C. 20001
(202) 220-1100
Jonathan.Kravis@mto.com
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.