Oral argument — Arista Networks, Inc. v. Cisco Systems, Inc.

Before we proceed with our arguments today, I'm going to turn it over to Judge Chen who has a motion. Yes, I do. Thank you, Chief. I'd like to make a motion to have someone admitted to our Bar. And specifically, I move the admission of Samuel Kenneth Witt, who is a member of the Bar and is in good standing with the highest court of the great state of California. I have knowledge of his credentials, and I'm satisfied that he possesses the necessary qualifications. Mr. Witt, although I did not know his middle name is Kenneth, has been my clerk for the past year. And I know him to be an excellent lawyer with not only a deep understanding of technology, but also the law. And he's very much helped me over and over again have a better, deeper understanding of not only the technology in a given case, but also some of the more intricate parts of the law. And I know if he hadn't been there for me, I might have made a few mistakes along the year. But luckily, I've made no mistakes during the year. So I thank Mr. Witt for that. And I also know him to be someone of really high character. And I'm very impressed with him on a lot of different levels, including his unusual lunch habits that make him his own man. And so I'm impressed with that, too. So I humbly move to have him admitted to our Bar. Letting the mistake matter go. It's a pleasure to grant motion on behalf of the panel. Please stand. Please let me by name. Do we solemnly swear and affirm that you will comport yourself with the attorney and counselor of this court, upright and abiding by the law, and that you will support the Constitution of the United States of America? I do. Welcome to the Bar of the United States Court of Appeals for Federal Service. Congratulations. The first case for argument this morning is 17-1525, Agrista v. Sisko. Ms. Degman, you're first. Oh, wait. So you're doing all of the original initial arguments? Yes, Your Honor. So I'll be handling Agrista's appeal, and Mr. Powers will handle the cross appeal. May it please the Court. We are a bit in an unusual situation where both parties agree that the board's claim construction was faulty because it excludes the disclosed embodiment, which describes broadcasting in the form of using multicasting, where a transmission is made to one or more devices connected to a network, but not, importantly, to all the devices. So you agree insofar as the board's construction was wrong, but you disagree. So why don't you explain to us the difference between your position and your friend's position with regard to the correct claim construction? Yes, Your Honor. I think what you'll note when you compare the construction side by side is that Sisko's construction adopts our construction and then tacks on some language at the end that the transmission occurs without specifying what devices will ultimately receive the data. And I think it would be useful to point out we all agree that the standard we're applying here is the broadest reasonable interpretation. And so there can be no dispute that Sisko's construction is more narrow than Agrista's construction. Sisko's construction attempts to adopt and read in wholesale the perverted embodiment, whereas Agrista's construction is broader and yet still reasonable because it comes directly from the language of the specification. And what's the difference, can you tell us in English, the difference between Agrista's construction and the board's construction? So the board's construction, yes, so Agrista's construction says transmission to one or more devices. Devices are, in fact, in these claims on the network. But the board's construction says that transmission has to be to each and every one of the devices on the network. And we should point out that these devices are described in the patent to be… I have a question about your claim construction. But before I get there, just so I don't get lost, when are we going to talk about the sign or estoppel? That's on the cross-appeal. That's the cross-appeal. That's the cross-appeal. I know that's the sexy issue. I'm sorry, it's not mine. Well, those are your words, not mine. Let's get, okay, back to claim construction. So I was making the point that… My question is, we're talking about claim 29 right now, right? Yes, Your Honor. And the independent claim talks about communicating. And now claim 29, the dependent claim, says communicate by way of broadcasting. So that suggests that communicating and broadcasting can't be identical. And communicate, I think, is essentially nothing more than communicating information to one or more devices. So that's why I'm wondering, why isn't broadcasting something more than just the mere act of sending information to one or more devices? I mean, it seems like under that theory of claim differentiation, it has to mean something more than just the mere delivery of information to a device somewhere on some network. So claim 1, I think you're right to focus on, it talks about communicating information about the change in configuration. It does not say to a device in communicating, as described in the patent at column 7… Are you suggesting that the word communicate could mean, like, walk into a closet and then communicate something out in the air and there's nobody that's on the receiving end? The patent tells us that the communication can occur by a blinking light on the device itself. So the communication can happen because there's an indicator light on the device itself that's at column 7 around line 25. So claim 1 is really quite broad. The communication is just letting someone know there's a change in configuration, which may not be a transmission over a network at all because it could just be the blinking light. And so what we see when we get to claim 29 is that it's a broadcast, which we know is over a network. But the key distinction to Your Honor's question is it's not to each and every device on the network because that would exclude this multicasting. Multicasting is a type of communication where there is an address, a multicast address, and people subscribe to that address. And the patent tells us that these network security monitors have to subscribe to the multicast address. And the patent doesn't say you must have two, and it does not say that each and every single device on the network has to subscribe to that multicast address. What if we just felt like flipping the term broadcast, which has an understood meaning in the art, to encompass everything including just merely sending an email from node A to node B is not right. It's just too inclusive. Is there a backup claim construction that you have? So I interpret your case law as not allowing me to come up with brand new constructions here on appeal the first time. Well, what if the panel is left with being unsatisfied with all the proper claim construction, whether it's the boards, whether it's Cisco, whether it's ERISA? And obviously we have some responsibility to land on some kind of claim construction. What would your backup be to sending an email from Joe to Tom? And so in the petition we suggested that kind of an ordinary meeting of the term was a transmission to, I think it was, all members of a group. So that's not everything on the network? All members of the group now? I guess your answer would really be you have to look at the specification and the usage of the term broadcast, and there's a clear notion in the specification, in your view, that whatever broadcast means it has to include multicast. That's absolutely right, Your Honor. And so at a minimum, a claim construction has to encompass perhaps the ordinary meaning of broadcast, because there was nothing in the specification that specifically displaced that ordinary meaning, but also be something inclusive enough that it encompasses the conception of multicasting. Yes, and it's important because the multicasting that's described in the patent, where there's an address that other devices subscribe to, is exactly the kind of broadcasting described in the Idiyama reference. And so from our perspective, that if you want more devices, a multicast address that's subscribed to is disclosed in the Idiyama reference, and there can be no reasonable dispute over whether that's the case. And so from our perspective, if you were to adopt a new construction under this court's cases, there really would be no need to ask the board to revisit whether Idiyama includes that, because Idiyama is squarely on all fours with any kind of definition that would be appropriate for the term multicasting. And a case came out after a brief Owens-Corning, 875 Fed 3rd, 896, that discusses situations where this court does not need to remand even when you change construction. And so under our construction, Idiyama definitely has it. I think under any construction that could stand scrutiny under this court's claim construction principles, Idiyama has broadcasting. So the last thing I'll touch on quickly is if there's no change in the construction, there's not substantial evidence to support the board's finding. I think the board did not really address and appreciate the second aspect. There's multiple embodiments within Idiyama. One is sending out an email to a group of devices, certainly, but it also talks about how the email can be too slow. There's a lag time. And so it talks about pushing out a notification such as a text message to the whole network, to the network. And that was cited by us in our papers to the board, argued to the board. You'll see in paragraph 55. Just real quickly, how do you define network? Because that seemed to be one of the puzzling questions below. Network could be an understanding that every and any single device that you're capable of being linked up and communicating with is part of a very large, broad network. But the network could also be subgroups that you create. And then so there's this private group of terminals that are able to communicate with each other. And creating that group, maybe it's an email group, email message group, that creation of that group is a network in and of itself. Obviously not as big as a broad general network, but maybe that's also a network. So I think that's a good question. That term was not up for construction before the board. It was given its ordinary meaning. And Judge Chan, I think you're quite right that under the broadest readable interpretation of network, it could be the Internet. It could be every possible thing that could possibly be connected, including my refrigerator. The claim doesn't say network, though. It doesn't, but when you read it. The claim construction says network. That's right. So now we're doing a construction of a construction. That's a very fair point. I think in the context of the 597 patent, we're talking about switches and routers that are connected to both the Internet, the wider Internet, as well as corporate networks and smaller networks. So to answer your question, I think under what would be the broadest reasonable interpretation of network, I mean, it would be both. I mean, I think a network can be considered everything, and it can be seen as smaller groups like you see in aspects of the Ileana reference. Okay. Why don't we move on to the other side? Thank you. Thank you, Chief Judge Prost, and may it please the Court. John O'Quinn on behalf of Cisco. While I'm happy to answer questions the panel might have with respect to the broadcast limitations, I'd actually like to start with the issue of a sign or estoppel because if the Court addresses that issue, that goes to all of the claims in the 597 patent and would resolve all of the issues with respect to the 597 patent. And a sign or estoppel, of course, is a substantive patent law doctrine that in the words of Justice Frankfurter, quote, has been part of the fabric of our law throughout the life of the nation. Well, before we get to the merits of a sign or estoppel, we have to get to the question of whether or not it's reviewable at all, right? Absolutely, Chief Judge Prost. And your view on that is? It is certainly an issue that is reviewable by this Court for the reasons explained by this Court in both Versada and credit acceptance. And indeed, I don't think that there's any principled distinction between the issues in this case and the issues that were in credit acceptance. To take a step back, as Versada explained, institution and invalidation are two distinct actions by the PTAB that do not become the same just because the agency decides certain issues at both stages of the process. And here, there's nothing about the issue of a sign or estoppel that is uniquely relevant to institution any more than the statutory estoppels that were at issue in credit acceptance, which this Court found governed at, quote, any stage of the proceedings as well as in the district court and in the ITC. And indeed, precisely because... Isn't Versada more about the power of the institution, the board, to undertake a review as opposed to a sign or estoppel, which is more about the eligibility of a given entity to pursue an IPR? So isn't there a distinction there? So Judge Chin, there is hypothetically a distinction there, and it's a distinction that this Court had relied on in Husky, but I think it's a distinction that no longer exists after Wi-Fi 1. In Wi-Fi 1, the time bar, of course, applied to an individual. And similarly, the statutory estoppel that was at issue in credit acceptance, it doesn't... By its own terms, if you look at 315E or 325E, those statutory estoppels are not a limit on the power of the board. They say that a petitioner may not pursue or maintain an IPR. And the exact same rationale would apply here to the doctrine of a sign or estoppel. To put it differently, precisely because a sign or estoppel is a substantive issue, whether or not it has relevance, whether or not it is something that the board considers, or the director, I should say, considers at the institution phase, does not render it irrelevant at the merits phase. And when you get to the merits phase, this Court can consider it just like you could consider the merits decision in Versada, just like you could consider the merits decision in credit acceptance. Husky, of course, by its own terms, was only addressing an institution decision. The only issue that was raised in Husky was whether or not the board was required or the director was required to apply a sign or estoppel at the institution phase. It was not raised subsequently. And in that sense, what was at issue there was only the very kind of preliminary decision that Wi-Fi 1 says is not reviewable, as opposed to the final decision at the merits phase. Did Husky say that these kinds of decisions on whether to apply a sign or estoppel are closely related to the decision to institute an enterprise review? Because it is closely related, therefore, it's not something that is reviewable. That decision on whether a sign or estoppel does or does not guard a petitioner is not something that's reviewable. So, Judge Chin, there were two parts to the analysis in Husky. The first was to ask the question, was the issue closely related to the institution decision? The second part was then this dichotomy between was it a limit on the power of the board or was it a limit on who could file? I think actually both parts of that analysis have now been overtaken by Wi-Fi 1. If you look at page 18 of the slip opinion in Wi-Fi 1, it specifically addresses what does the closely related to idea mean? And what the Supreme Court says in Cuso about closely related is the question is whether or not the argument being advanced by the patent owner depends on something that is closely related. We're not relying on 311. We're not saying the board violated 311. Just like in Cuso, the argument by the patent owner was that the board had violated 312. Our argument is post-institution, the board was required to apply this on the merits, under 316E, for example, in determining whether or not the petitioner could meet its burden as to patentability. Now we're getting into the merits of the argument, which is fine. I'm having a hard time seeing what that has to do with this matter. Why that statutory provision, that deals with who has the burden on patentability. That seems to be a separate and distinct question to me. My point in referencing 316, Chief Judge Prost, was just simply to note that there are any number of reasons that a party may not be able to bear, to meet its burden. For example, the application of collateral estoppel may be a reason that a party cannot bear its burden. The only point that I was trying to make is that this is something that the board, just like the statutory estoppel at issue in credit acceptance, that the board has to apply in rendering its final written decision. Even if you think that it is relevant to the 311A question, it is not uniquely relevant to the 311A question. That actually brings us to the merits of the assigner estoppel issue. The board offers two interrelated reasons for refusing to apply assigner estoppel. It first says that when Congress wants an equitable defense to be applied, Congress will say so expressly. Then it points to other PTAB decisions relying on Section 311A, which 311A merely provides that a person who is not the owner of a patent may file an IPR petition. Now, neither of these justifications can explain casting away a 140-year-old doctrine of substantive patent law. And the first point I think is the most important. The doctrine that the board never applied in the inter-parties re-exam context. Well, to my knowledge, Chief Judge Prost, the only time that that was before the board was once in a non-precedential unpublished decision by the board And that is certainly not a substantive gloss on the law that Congress is presumed to adopt. I mean, if you look at the Supreme Court's decision in JAMA versus ICE, the Supreme Court's clear that even a couple of court of appeals decisions is not enough to provide that kind of gloss. You know, what I'm trying to figure out in looking at the structure of the statute, we have 311, which tells us who is eligible to pursue an IPR. That gives us one way to read it. And then we have 315A and 315B and 315E, which tell us identify categories of people who are not allowed to pursue an IPR. And then we have the Trademark Act and the ITC statute, which both say in those particular administrative proceedings, here in the statute, we're telling you that a party can advance equitable defenses. And here for the AIA, with respect to these patent office proceedings, there's no such corresponding statutory provision. And so I'm trying to figure out, isn't there an inference there that we should be considering, given that the statute for the patent board is silent on this? There was no historical evidence that the patent office ever applied a sign or estoppel. And then these other places, Congress specifically said, yes, you ITC, consider equitable defenses. Yes, you trademark board, you consider equitable defenses. You patent board, I'm not saying anything about. So what am I supposed to take away from all that? Sure. No, thank you, Judge Chin. And I think your question packs in probably most of the points that I want to try to address in the time that I've got left. So let me break it down. First, I think underlying your question is, what's the starting point? What is the background principle? When Congress doesn't specifically speak one way or the other, what is it that you are to assume? The Supreme Court has answered that question time and time again. The common law principle is well-established. Courts may take it as a given that Congress has legislated with an expectation that that principle will apply. And it reiterated that in B&B hardware, which is directly relevant here, where the court said that you could take it as a given, that Congress had legislated with the expectation the principle of collateral estoppel would apply. And that's true not just with respect to courts. Just so I understand your argument, the Trademark Act did not need to have that provision about equitable defenses being available in contested proceedings in front of the trademark board. Nor did the ITC statute need to say that part about equitable defenses being available for those to be applied because there's just some backdrop principle where in these administrative proceedings, of course, you can always apply these equitable defenses. That's absolutely right, Judge Chin. It's a little superfluous, then, that those statutes say what they say. I would say that it is belt and suspenders that those statutes say what they say. And indeed, if you look at the ITC statute and you look at the underlying decision in the Intel versus ITC matter, namely the decision by the commission itself, the commission wasn't saying, well, we can apply a sign or estoppel because the statute says this. It recognized that actually by its terms, the defense, a sign or estoppel wasn't a defense in the sense that the statute was referring to. It was just saying, well, this is yet another data point that reaffirms that Congress intended background principles to apply. And I think if you look at this court's decision in Rios versus Nicholson, which involved application to an Article I court, which is effectively an administrative agency for these purposes, there you had the issue about the postmark rule and whether or not that was compatible with the mailbox rule. And the court came to the conclusion that you continue to apply the background common law presumption unless it is incompatible with the statute itself. And there is nothing here that's incompatible. When you look at what 311A is about, it is about whether or not you can proceed ex parte or whether you have to have standing. That's what it's about. And a petitioner doesn't have to be a party who has standing. That's actually in contrast to the trademark statute, a different part than what you were referring to, Judge Chin. But if you look at 15 U.S.C. 1064, which governs who can file trademark cancellation proceedings, it says that those can be filed by a person who can demonstrate that they would be injured by the continued registration of the trademark. Congress didn't draw that line here. Congress said you don't have to have standing in the sense that you don't have to show that you would be able to be injured. And at the end of the day, that's what 311 is all about. It is about who can file a petition. It's not about whether or not you can ultimately prevail on your petition. And there may be any number of reasons, whether it's because of the law of obviousness or anticipation or whether it's because of collateral estoppel, which the board or the patent office applies in other contexts. And that's another point here, that the patent office has taken an inconsistent approach. Time and time again, this court has recognized that, as a general matter, the patent office has the power to apply equitable and non-statutory doctrines. That's what this court's decision in In re Bogues was all about, saying that we're not going to deny to the patent office the power that a district court has. And I don't think that you can look at the specific prescriptions in 315 and say, well, that is Congress answering the question. The things that you're talking about in 315 uniquely arise in the context of IPRs. Congress had to create them in order for them to exist. Otherwise, there wouldn't have been a time bar. There wouldn't have been an estoppel if you had already filed an IPR. Congress didn't have to create a sign or estoppel. It had been there for 140 years, just like Congress didn't have to expressly say. 315B is a classic form of litigation estoppel. I mean, it seems like if it didn't exist, we would still use it. But 315B is different from normal collateral estoppel. 315B prohibits you and your privies from proceeding if you've been served with a complaint, regardless of the resolution of that case. If you've been served with a complaint and one year goes by, you can't file an IPR. That's not the common law. That's not the way that the common law would work with respect to either raised judicata or collateral estoppel. Now, there may be some common law abstention principles that potentially get applied, but those are— I know you're running out of time, but let me ask you a quick question about— There's—the statute has a gap, and now the Patent Board has stepped in and issued a presidential opinion where there's 275 board judges that have stepped into that gap, and it's been endorsed by the director of the agency. Do they get Chevron deference? That suggests that could have kinds of hallmarks of granting Chevron deference. Why not give Chevron deference to the presidential board? Sure. Well, first, Judge Tinn, the only presidential board decision that is in this space is the Athena decision, and by its terms, it is limited only to the issue of institution, not whether or not the board applies a sign or estoppel when you get to the merits phase. So even if you thought that Chevron deference would apply to a presidential board decision, you do not have a presidential board decision addressing the issue that is before the court, which is whether or not the board is required, not when it's acting as the designee of the director in instituting. And there are any number of policy reasons that the director might decide, I don't want to consider this at the institution phase because maybe the evidence isn't going to be able to be marshaled. It's not the kind of thing that the petitioner is going to usually be in a position to prebut. They don't get a response before the institution decision to marshal the evidence and decide whether or not a sign or estoppel would apply. That's an institution issue, and the Athena decision, Athena v. Husky, only addresses institution. The non-presidential decisions that are at issue in this case and in the following case, of course, I think this court, a majority in Aqua Products, agreed that a non-presidential board decision does not get Chevron deference. If you were to think that the board's decision was, in Athena, was relevant here, I would submit there are two reasons it should not get deference, and the court's review should be de novo. Number one is it is not a reasonable interpretation of law. Again, the Supreme Court has made clear you have to, when they legislate, it's with certain background principles in mind, and the board flips that on its head. It's just wrong. And second, what the board purports to do in interpreting Section 311A, it's not trying to fill a gap in a policy perspective. It's applying principles of statutory interpretation, and as the en banc D.C. Circuit held in Atkins v. FEC, if what you're doing is just interpreting judicial precedent, that's not entitled to Chevron deference. Okay, your time is way off, but if you want to take two minutes, no more, and talk about broadcast to claim construction issue, we'll allow that. Thank you, Chief Judge Prost. Just two very quick points with the court's indulgence. The first is Arista's proposed claim construction would be an unreasonable reading of the term broadcast. It would encompass the hypothetical that Judge Chin was asking about, where person A sends an email to person B. That cannot possibly be the meaning of the term broadcast. Now, the definition that we had advocated before the board was one that would encompass the concept of multicast. That is, that you were sending it out, but you do not know, to one or more, but you do not know who the recipients are going to be. And I think that's sort of the key issue, is that you don't know who the recipients are going to be. Now, I don't think that you get to this issue not only because of the assign or estoppel hurdle, but also because I think that there is an institution issue in terms of how it was that Arista pitched its petition and the alternative arguments it made to the board. Arista didn't advocate that the board invalidate based on Iwayama, based on the construction that they were advocating. That was an alternative argument. And the last point that I'd make on this, Your Honor, is that to the extent that you think that you do reach the board's construction and that the board's construction is wrong, even if you were to adopt Arista's construction, which you should not, that would still have to result in a remand, because at the end of the day, I think that Iwayama is using the term network in a very different way than it's being used in the patent. Thank you. Do you think Iwayama discloses a form of multicasting? I don't think that it does, Judge Chinn. I think it's ultimately about group chats, and I don't think that that's about multicasting to network devices in the way that's contemplated by this patent. Thank you. We'll restore some new data. Thank you, Chief Judge Proust. Thank you, Your Honor. I'd like to begin by addressing there's three specific issues with regard to Assign-a-Restoppel, and I'd like to take them in order with the Court's permission. The first is reviewability. The second is whether the statute is clear that the board's construction of its ability to eliminate Assign-a-Restoppel is correct, and the third is the Chevron deference issue. With regard to reviewability, the key question I think is the one that Judge Chinn posed, which isn't Versada and Wi-Fi One and Quozo, all dependent on the question of whether you're enforcing a limit on the statutory authority of the board to act. That was true in Versada as to whether it was a covered CBM patent, and it was true specifically in Wi-Fi One because the board did not have statutory authority to act. And the response from Counsel for Cisco on that particular question was that Wi-Fi One eliminated the issue of whether it's relying on authority to act, and that is squarely inconsistent with the very last paragraph of Wi-Fi One explaining its reasoning, and it's important enough that I'd like to read it. Quote, Enforcing statutory limits on an agency's authority to act is precisely the type of issue that courts have historically reviewed, citing cases. As a statutory limit on the director's ability to institute IPR, the 315B time bar is such an issue. We hold that time bar determinations under 315B are revealed by this court. That is the very last concluding sentence of the opinion before the conclusion section. That explains the reasoning of the Wi-Fi One court, and to say that Wi-Fi One eliminated the distinction between enforcing and reviewing statutory limits on the agency's ability to act and other things is simply, I think, not an accurate reading of Wi-Fi One. And when you look at what the agency's decision on the Sino-Rostovl is, clearly not a statutory limit on the agency's ability to act. It is a discretionary decision as to what defenses that agency will entertain in its review conducting its job, and it's consistent with what Clozo and Wi-Fi One highlighted as being the purpose of IPRs, which are to address and eliminate patents that should not be in the system. What if there was no such thing as Section 315 in the Act? No 315A, B, or E. And then the petitioner was someone who had litigated the case as a defendant who challenged the patent and lost, and it was a final decision. The patent's invalid. I mean, the patent's valid. And then after that final decision, they try to challenge the patent again through an IPR, and there's nothing in the statute that actually says there's a bar to that And the patent owner says, wait a second, you can't let this person in because they've already had their day in court. And then the board said, we don't have to consider that. We don't have to consider that common law doctrine. Do you think the board would be right or wrong in saying that? On the question of reviewability, I don't think it would be a statutory limit on the board's ability to act, and therefore it would fall on the other side of the line. On the question of whether that proceeding would be allowed to go forward in light of the history that you've described, I think then you're sort of getting into a Fresenius question. Because the court has addressed that issue separately as to which proceeding trumps when you have two proceedings. But on the question of whether the board The question is, does the party who lost in that prior litigation on trying to knock down the patent, should they be allowed to invoke this proceeding in front of the agency, basically having a do-over? I think that's A literal second bite. Absent something in the statute saying no, and absent the application of a doctrine that makes sense to say no, then they would have an opportunity at the bite. The question more relevant here would be, if that agency invoked some other doctrine, not statutory limit on its ability to act, to deny that person access to that proceeding, would that be reviewable? I think the answer is no. Because that then would be that agency's determination of how it conducts its business, not whether it's permitted to conduct its business. And those decisions are not reviewable. Why don't you move on to part two before you run out of time? Because you've got three parts. Thank you, Your Honor. So part two is a question of whether this, the agency's interpretation of its authority is proper. 311 is clear. It says a person who is not an owner may file. But what about your friend's foremost argument, which is this was written in the backdrop of hundreds of years of principles and that governs? I agree completely that it's written with the backdrop, but I disagree with the conclusion. The backdrop is this. That same language effectively was in pre-AIA, inter partes re-exams. That was the test about who could approach the patent office for that particular proceeding. Wait a second. The sign in Aristotle was never in the inter partes re-exam statute. No, I'm saying the language about who may proceed a person who's not the owner. So that same language existed pre-AIA and pre-AIA, the patent office interpreted that language to preclude the use of the sign of Aristotle in pre-AIA inter partes re-exam proceedings. And importantly, the important piece of that is that as the Supreme Court's case in Lorillard makes clear, Congress is presumed to know about that pre-AIA interpretation by the PTO of that same language. But your friend's answer to that, I think, at least in part, is that they only did that once. Not true. There's once in 2007 for sure, but you don't need more than once. There's also, as we've cited, Mr. Chisholm's treatise makes clear, so it's not like it's hidden. And Mr. Chisholm's treatise cites a 1997 decision as well. Well, what if we were to say this one decision, non-presidential, not enshrined in the manual or regulations or anywhere else, is just simply too obscure and too slender of a read for us to assume that Congress was well aware of it and wrote the AIA against that lone, random-ish decision? So then where else do you have... I don't believe that would be consistent with Lorillard, because I don't think Lorillard is dependent on there being more than one decision or something else, but... The agency issues thousands and thousands and thousands of decisions every year. And I don't think... It's the tiniest of needles in that haystack. True. I don't think Lorillard depends on an actual fact of knowledge by the Congress. I think it's presumption. But I take your point. Then I think the important fact is, and this was relied upon by the Court of Husky, is that the PTAB, unlike the ITC, does not have a statutory provision saying all equitable defenses shall be applicable. That's conscious. That's not random. The ITC specifically has something saying all equitable defenses are applicable. That's why a sign or estoppel applies in the ITC. That same provision does not exist in the PTAB. And you cannot assume that that omission was negligent. That omission has meaning. And therefore, it is certainly reasonable for the PTAB to say, as it did in Husky, that a sign or estoppel as an equitable defense is one we are not going to apply, consistent with our objective, our statutory objective, which is to clean up bad patents. And that statutory objective was relied on heavily by the Supreme Court in Closo as being a basis for its decision. The PTAB has that mission. And if it decides that... So there's nothing in the Patent Act that says anything about the Patent Office having the power to use prosecution legis or obviousness type double patenting or inequitable conduct or other doctrines that are rooted in case law outside of the statute. Is it your view that the agency shouldn't be using any of those doctrines? It's my view that the agency is free to decide which of those doctrines is consistent with its statutory purpose. And the statutory purpose is to clean up bad patents. And if it says, I have a petition in front of me that makes clear this is a bad patent, I am not going to stay my hand merely because of an equitable doctrine that we've said shouldn't be applied. I'm going to follow my statutory mission to clean up bad patents. And those other doctrines may or may not be as related to that statute. That sounds more of a deference question, discretionary question. We're trying to figure out what does the statute permit or not permit the agency to do. The statute clearly does not prohibit the agency on its face from deciding how it's going to conduct its business. And the statute's omission of a directive that all equitable defenses shall be applied is huge. Wait, so is it your view that it's up to the PTO, given the statutory, given 311, the PTO can go either way and it's decided, as Judge Chen said, it sounds like a deference thing. Or is it your view that the statutory language is so clear that this is it? On a sign of restoppel, I believe the statutory language is so clear. Any person, not an owner, may file. So your view is that if the patent office decided for whatever reason that they should follow what the courts have done, that would be contrary to the statute? It would be. It would be contrary to the statute because it's not obligated to apply all equitable defenses, as the ITC is. It is obligated by statute to allow anyone, not an owner, to file. And if you apply the doctrine of a sign of restoppel and do not permit a signer to file, then you are violating your statutory authority. Okay. Part three, deference. Part three is deference. So I heard three arguments made with regard to deference, but I think it's quite clear that this is within the agency's authority to act. So there's really no question that this is something where Chevron deference would normally be accorded. The first argument that was made was that... What are we deferring to here? We're talking about... The presidential decision in Husky. Okay. And you disagree with your friend that that really went to this issue? I do. You read that decision. It happened to be in the context of a petition to institute, but the reasoning was not limited to that in any way. Okay. The reasoning says our statute says anyone, not an owner, may file, and therefore a sign of restoppel should not be applied. It wasn't saying it could be applied at the merits phase. That would be silly. You don't apply it at the institution phase, but do apply it in the merits phase? That would certainly not be a fair reading of that decision. So I think that issue is easily resolved. The second was it's not a reasonable interpretation. Well, I disagree completely with that. It's reasonable because of the fact that the PTAB doesn't have the same language as the ITC about all equitable defenses. It's reasonable in the sense that there was a PTAB decision in the pre-AIA language saying we interpret that to preclude an application of a sign of restoppel. Whether it was a needle in a haystack or not, that is a fact. It was reflected in Chisholm. It was reflected in a district court decision as well. And to say that those facts should just be ignored because it's only a single decision I think is not fair to the PTAB. They've made that decision. Have you read our package of aqua products opinions? I have. Okay. And what do you think? Do you think those opinions collectively foreclose the view that the board, by issuing a presidential opinion, can get Chevron deference for that presidential opinion? Assuming there is some gap or ambiguity in the statute and then they attempt to explain why they make one choice over the other in the face of the statute. Two thoughts on that, Judge Shen. I've read them carefully because they're obviously directly relevant to this question. They're addressing a slightly different question. They touch the relevant, of course, but they do not address directly the question of whether a presidential decision by the PTAB would or would not be the type of action that would be entitled to deference. And if they were interpreted to do that, that would be squarely inconsistent with the Supreme Court's decision in Mead. Mead says unambiguously that you do not need formal rule and comment decision making and, in fact, specifically says in footnote 12, I believe, that formal adjudications can be the basis for Chevron deference. There are three examples that they give in footnote 12 and in the text. One is the nation's bank case. That was merely a letter from the controller. It wasn't rule making, decision making. It was simply a letter from the controller. That was given deference. Chevron type deference? Yes. The INS case versus Aguirre, which is 526 U.S. 415, also cited in footnote 12 of Mead, was simply a prior Bureau of Immigration Agencies decision, a court decision. The National Railroad case, also in footnote 12 in Mead, also just a prior decision. And so I don't think we can read the aquaproducts cases to go so far as to say that you do not give Chevron deference to a prior precedential decision. And let's be clear, PTAB decisions aren't routinely granted precedential status. There's a whole process for that, as Judge Chen knows very well. Not very many are accorded that status. This one was accorded that status in 2017, four years after the decision was rendered, so it was very significant that it was done so. And the idea that that decision, with that imprimatur of the entire agency, would not be given Chevron deference I think is squarely inappropriate, and certainly foreclosed by Mead. And so I think there's not really an argument that you would not give Chevron deference to the agency's interpretation of its ability to act and how it's conducting its business. Okay, thank you. Thank you. Judge Press, you're very kind. I will try to be under a minute. The only thing I wanted to address that was brought up was this notion that somehow we can't argue that Iwayama discloses broadcasting if you were to adopt our construction. And I would submit respectfully that throughout the proceedings and the petition and the reply and the oral argument, there could be no question that we were arguing that Iwayama met the broadcasting limitation in the combination. And did we use the specific words, it has one or more devices? We did not, I will tell you. Those three words do not appear. But we argued it met the limitation under much more narrow constructions. And by saying it met the one or more groups of terminals or even under Cisco's construction, which was exactly ours with that material tacked onto it, we said it met that construction as well. Everyone was on notice. There can be no gotcha here in terms of our arguments consistently that Iwayama discloses broadcasting under our construction, their construction, the board's construction, the construction you may think on your own, Iwayama has it. So it's not me. Thank you, Chief Judge Prost. Responding to the points that Mr. Powers made, more or less in order. First, if a sign or a stopper were viewed as being purely an institution issue, they would create the anomaly of, number one, stripping Cisco of what would otherwise be a complete substantive defense to patent invalidity against ARISTA in any other forum, in district court, in the ITC, in any other forum, and doing so without any judicial review, all without Congress expressly saying so. That is fundamentally inconsistent with the, quote, long tradition of judicial review of government actions that alter the rights of an affected person, as this court said in Versada. Now, when my friend was arguing and saying that Wi-Fi One and Versada both just go to the board's power, he left credit acceptance off that list. Credit acceptance is indistinguishable from what you have before you. It's the same type of a stopper. Yes, it's in the statute, but by its terms, the way it's written, it's not written as a limitation on the board's power. It's written as a limitation on the person. And indeed, Wi-Fi One did reject the notion of the who versus what. The panel decision in Wi-Fi One, Akati's, those had been viewed as a limitation on who, and that's what Husky relied on. That's gone after Wi-Fi One. And, in fact, I think at page 18 of the slip opinion of Wi-Fi One, the court makes clear what it has in mind as far as what is likely unreviewable. It's the unreviewability being, quote, limited to the director's determinations closely related to the preliminary patentability determination or the exercise of discretion not to institute, and it recognizes that the preliminary patentability determination is not reviewable because the final patentability determination is, and that's where we're at. We're at the failure to apply a sign of restockable in the final patentability determination. Now, with respect to the issue of deference, I think it is very clear, Judge Chin, that at least a plurality of this court in ACWA products critiques the idea of relying on even, of giving Chevron deference even to precedential board decisions. That plurality relies on Mead. There's the Insignia-Motocar's case, and it would be strange to say that the board would effectively get deference in deciding what is a substantive issue of patent law. That would be inconsistent with this court's observations in Cuso at 793 F3rd at 1279 that while through notice and comment rulemaking it can do procedural things, it can't shape substantive patent law, and that's a point that this court made looking at an earlier statute in Cooper versus Dudas at 536 F3rd at 1336. Mead recognizes, look at what it is that Congress contemplated that the agency would do, and what Congress contemplated is that if the agency wanted deference, it could do so through notice and comment rulemaking. So I think it would be inappropriate to defer to the board's decision here, but in any event, even if you do, it's ultimately unreasonable, and the reason it's unreasonable is twofold. Mr. Powers relies, again, on this unpublished, obscure decision in Henry Harvey. I don't think that given the Supreme Court's decision in JAMA versus ICE, which says court of appeals decisions can be too obscure to recognize that there's a consensus on what the language means, that you can rely on that. And even if you could, Congress was executing a fundamental shift in going from inter partes review. It was moving from an examinational system to an adjudicatory adversarial system. And signer estoppel, like collateral estoppel, is the kind of thing that applies in these adversarial proceedings. And that brings me to my last point. Mr. Powers suggests that collateral estoppel itself shouldn't apply. That's a shocking proposition given how fundamental that is in the case law. It is inconsistent with what the Supreme Court said in B&B hardware, and it is inconsistent with what the Patent Office does in every other context, where it applies things like collateral estoppel, judicial estoppel, and notions of even prosecution history, which are effectively a form of estoppel that the board applies in doing claim construction in these cases. So a signer estoppel should apply. It should bar Arista's arguments, and the decisions with respect to the 597 should be reversed. I'm happy to answer any questions you may have. Thank you. Thank you, Chief Judge Crost. All right.