[00:00:00] Speaker 01:
Our next case is Video Labs v. Netflix, 2025, 1131. Mr. Cardenas, Navia.

[00:00:11] Speaker 05:
Good morning, Your Honors. May it please the Court, there are three independent bases to reverse the Board's finding that the 790 patent is invalid. Two of those turn on issues of plane construction. The third is based on a failure in the petition to analyze a limitation. All are subject to de novo review. I'll take them in turn. The first issue is whether the claims require a stored association between an item of content and multiple implementations of that item of content. That's a one-to-many stored relationship.

[00:00:43] Speaker 05:
And the board found that that is not a requirement of the claims. Video Labs contends that it is.

[00:00:49] Speaker 05:
And the support for our claim interpretation comes directly from the plain language of the claims themselves. I'll refer to Claim 2, which everyone agrees is representative of the claims in the patent. And the first claim limitation requires, among other things, receiving and storing multiple implementations for an item of content. The next limitation then requires associating that item of content to each of its implementations in a product catalog. And so if you combine those two descriptions from those two claim limitations, you arrive at this requirement of storing a one-to-many storage of items to its implementations, to all its implementations.

[00:01:30] Speaker 05:
And that's – and the court found that wasn't a requirement of the claim and on that basis found that Meta anticipated this claim.

[00:01:40] Speaker 02:
But in fact, everyone – Does that really require some sort of hierarchical arrangement?

[00:01:47] Speaker 05:
So we do use the term hierarchical arrangement, and by those terms, all we mean is what I just laid out as an express requirement of the claims. And the reason for that is because if you have, as shown in Figures 5 and 6 of the 790 patent, if you have your items… and then they're each associated with one or many different implementations, you can diagram that as your items up here and then connections linking them down to their various implementations in a tree-like structure or in a two-stage structure.

[00:02:18] Speaker 05:
So by using that term hierarchical, we weren't adding anything beyond what is already in the claims. It was just a shorthand to refer to those claim requirements.

[00:02:26] Speaker 02:
But the claim doesn't require some tree-like structure, does it?

[00:02:31] Speaker 05:
No, Your Honor. It's just a way to visualize it, a way to diagram it.

[00:02:35] Speaker 02:
It just says there are multiple implementations of at least one item, and then ultimately only a single one is displayed.

[00:02:49] Speaker 05:
That's correct. And it's that the multiple implementations have to be stored, have to be linked to the item of content in the product catalog. And that is, I think the board maybe read in more into what we meant by hierarchical. All we meant is that requirement of a one-to-many linked relationship in the product catalog.

[00:03:12] Speaker 05:
And on the board's factual findings, it's undisputed that Meta does not do that. The way Meta operates is it stores applications in its database, and in each application with that database always links just a single item of content, which is referred to by its title, to a single implementation, which is referenced by the URL in that application. And so in Meta, there is no way, there is no association between of an item of content to multiple applications.

[00:03:43] Speaker 05:
Maybe that could have been done if each application could have had multiple URLs or if there was some linkage between the applications that are disclosed in meta. But everyone agrees meta doesn't teach that. And the board's decision did not hinge on that at all.

[00:04:00] Speaker 05:
And so really the board's decision hinged on this claim construction issue of Do the claims require a one-to-one relationship stored in the product catalog, or must it be a one-to-many? And our position is a one-to-many that's consistent with the embodiments disclosed in the claims. We were not reading anything from them into the claims. They're just indicative of the type of relationships that are required to satisfy the claim language.

[00:04:30] Speaker 05:
Unless your honors have any question about that, I will turn to the second issue.

[00:04:36] Speaker 05:
The second issue is about the construction of the presenting claim limitation. And here, Video Labs' position is that this limitation requires an affirmative act of filtering out any additional compatible implementations that might exist for a given item of content. And that contrasts with how the board interpreted the claim, which was to say, let's not focus on the process. Let's not focus on how you arrive at a presentation of just a single description of each item of content.

[00:05:11] Speaker 05:
Let's just look at the result. And if we see that at the end of the day you're presented with just a single description for each item of content, that's sufficient to meet this claim limitation. And we disagree with that. The problem with that claim interpretation is that it reads out the final clause of the presenting step, the requirement of presenting the single description regardless of the number of implementations of each said clause. item. And we can see that because under the board's view of the world, it's completely irrelevant how many other compatible implementations there must be.

[00:05:46] Speaker 05:
Again, all you care about is the results. The board, in a sense, read this final claim limitation as almost an if-then of almost if there are multiple compatible implementations, then you need to perform this second filtering step. And it found that that didn't apply to meta because it found it as a factual determination and decided that meta never has multiple compatible implementations. And it said that's okay because the claims don't require that. You only need to do that second act of filtering if you have multiple compatible implementations, which isn't present here.

[00:06:26] Speaker 05:
And maybe this will be helpful to clarify the dispute. An analogy here is you can imagine law firms might have a checklist of tasks to perform before a brief is filed with the court, right? And one of those could be take a look at all the citations in the brief and make sure they are accurate regardless of the number of citations that are in the brief. And what that tells us is if you wanted to see and assess whether that task on the checklist was performed, you can't just look at the end product brief and say, great, all the citations are correct.

[00:07:02] Speaker 05:
Because, again, to perform that task, you needed to actually undertake the step of reviewing the breadth, the brief, and making sure that all the citations are accurate. And again, it's regardless of the number of citations in the brief. The same is true here. It tells us that this presenting step, the second act of filtering, is performed regardless of the number of implementations for that item of content.

[00:07:31] Speaker 05:
Unless your honors have any question about that issue, I'll turn to the third issue.

[00:07:38] Speaker 05:
The third issue is that in Netflix's petition, again, for this same limitation, this regardless limitation, it just was not addressed in the petition. And under Magnum Oil and Core Photonics, that's not an issue that can be remedied in Netflix's reply brief. But in fact, that's what happened. Netflix ended up substantively addressing that limitation for the first time in its reply brief, and the board then credited Netflix's analysis in the reply brief in its final written decision.

[00:08:10] Speaker 05:
And that's not proper. In our view, the only relevance the reply brief has and what Netflix said in it is to highlight the gap between what was in the petition and what was in the reply brief.

[00:08:22] Speaker 05:
In the reply brief, Netflix, for the first time, analyzed figures 2 and 11F of meta and explained how they supposedly met the presenting limitation. The petition, if you look at where this claim limitation is very generally analyzed, there's no discussion or analysis of figure 2 or 11F. And at that point, again, this issue is reviewed de novo. If this court agrees that Netflix did not sufficiently analyze this limitation in its petition, it's not something that can be cured at this stage.

[00:08:57] Speaker 05:
And if your honors have no further questions about that issue, I will turn to the final issue.

[00:09:03] Speaker 05:
For both of these issues having to do with the presenting claim limitation, that finding of anticipation by meta was backstopped by a finding of obviousness based on the combination of meta with a reference schlapper.

[00:09:19] Speaker 05:
But the challenge with that is if this court agrees that the presenting limitation was not disclosed by META, it's not anticipated by META, then that finding that it does was necessary to the board's obviousness analysis. And so at the least, at the very least, if the court agrees with us on the presenting limitation not being disclosed in META, The obviousness findings also need to be vacated. And we believe on this record, it's clear enough that this court should move forward and reverse the finding of obviousness.

[00:09:52] Speaker 05:
And there's two primary reasons for that. Again, one, under the board's fact finding, the meta reference never has situations where it ever has multiple compatible implementations of the item of content.

[00:10:07] Speaker 05:
And based on that, there would be no reason to modify meta to perform the second act of filtering out multiple compatible implementations. It's just never a problem that arises. So what would be the rationale to take that next step? And on top of that, again, if Meta doesn't disclose this regardless limitation, then Netflix hasn't identified any reference in the prior art that discloses the regardless limitation. And so there's nothing to add to Meta to even get at the act of that second filtering step.

[00:10:41] Speaker 05:
And it looks like I'm about out of time. So unless your honors have any questions, I'll save it for rebuttal.

[00:10:46] Speaker 01:
We will save it for you.

[00:10:48] Speaker 05:
Thank you.

[00:10:49] Speaker 01:
Ms. Carano.

[00:11:02] Speaker 00:
May it please the court, Elisa George Carano of Wilkie, Farr, and Gallagher here on behalf of Netflix.

[00:11:08] Speaker 00:
Now, I'd like to talk about that first issue that counsel for video labs addressed because he rushed through the claim language. And I'd like to walk you through it because if we look at the plain language of the claim, That confirms the board's finding in association with issue. So the claims require receiving and storing a plurality of items of digital content. And it also requires a plurality of different implementations of at least one of the items of digital content.

[00:11:41] Speaker 00:
And Video Labs does not dispute that the board's findings that meta teaches both of those limitations. The claims then require that the server system maintains a product catalog. With respect to the product catalog, all that's required is that the product catalog include, in association with each item of digital content, a reference to each implementation of said item of digital content. And that's all that the product catalog requires.

[00:12:14] Speaker 00:
What this means is that the claims were drafted broadly enough to allow for different database structures to fall within the scope of the claims. Because all that's required is that the product catalog store, that the reference to each implementation of said item of digital content be associated with that item of digital content. There's no other requirements for the product catalog. There's no language specifying how the product catalog should be structured.

[00:12:45] Speaker 00:
There's no language specifying how that relationship between the references to each implementation and its respective item of digital content should be stored. In other words, there's no limitation on how that association be stored. So there's no requirement.

[00:13:04] Speaker 03:
How does Meta meet this association of referencing different implementations of the same product?

[00:13:13] Speaker 00:
So the board found, and there's substantial evidence to support this, that in META's MOS system, when there's, and I think the best way to explain this is through example. And the example that's been below and discussed in the briefs here is when The item of digital content is Microsoft Word. So what Meta's MOS system would do is it would store one application profile with the title Microsoft Word. And then it will have a URL to the implementation for Microsoft Word for Windows.

[00:13:46] Speaker 00:
And then it would have a second application profile. And that application profile would have the title Microsoft Word. And then it would have the URL to the Mac version of of Microsoft Word. And both applications would be stored in the MAS's data repository, which maps to the product catalog. And both URLs are references to the implementations of Microsoft Word. And both are associated with Microsoft Word because the item of digital content, because they both have the same title, Microsoft Word.

[00:14:24] Speaker 00:
The title Microsoft Word represents the singular item of digital content, and there's substantial evidence for the board's finding on that. And what Video Labs has argued now is that...

[00:14:39] Speaker 00:
that there's this requirement of like a one-to-many in the product catalog. And that requirement is just not there. There's no requirement that the different implementations be clustered together in a hierarchy around the item of digital content. And there's certainly no requirement that the product catalog include product entries. So long, and there's no requirement that the product entries be structured in a certain way, so long as that relationship between the different implementations of item of digital content are associated with the respective item of digital content.

[00:15:14] Speaker 00:
And so if the patent...

[00:15:17] Speaker 00:
wanted to limit the claims to what's depicted in figures 5 and 6 of the 790 patent, it would have described that in a more, and it would have claimed that in a more expressed way in the claims, but it didn't. And the patentee knew how to describe those embodiments depicted in figures 5 and 6, because it described that in the specification, and yet that language requiring product entries is not in the claim. It didn't make it to the claim. And the specification is consistent that the claims are not limited to Figures 5 and 6 for the product catalog because the language that the patentee used is permissive.

[00:15:57] Speaker 00:
Just shorthand summary what this specification says. It says that the product catalog can be structured this way. It doesn't have to be, and it's not a requirement that the product catalog must be structured this way. And so even though Figures 5 and 6 are the only depictions and descriptions of how the product catalog can be structured, The claims, it's not limited to that one embodiment. In fact, the patent says expressly that the invention is not limited to the embodiments described.

[00:16:33] Speaker 00:
And so this argument, there's actually, it's important to get some context as to how this argument was raised because below, video labs did not say that this was a claim construction argument. Even in its sore reply, when it responded to Netflix's argument saying this is not a claim construction argument, and it said that in footnote four, which is appendix 2430, it said that all we're doing is responding, is analyzing the prior art to the claims. So even though this was belatedly raised by Video Labs as a claim construction argument during oral argument, the board still considered what Video Labs was arguing with respect to claim construction, and it disagreed.

[00:17:18] Speaker 00:
It put forth a well-reasoned analysis and said that the claims are not required to be limited to the figures in Figures 5 and Figures 6.

[00:17:30] Speaker 00:
because doing so would improperly import a preferred embodiment into the claims, because there's no clear language that the claims be so limited. And so now, for the first time on appeal, Video Labs is now arguing that, well, if this panel agrees with the board, then it's going to render this in association language with superfluous. And it's doing no such thing, because If the in association with language is removed from the claims, then all that would be required in the product catalog is that you would have names of items of content and all these URLs that are references to the different implementations, but there would be no connection between the two.

[00:18:13] Speaker 00:
And that's what that in association with language requires, that there be a way to map the the different implementations to the respective item of digital content. And that's what meta does, as I've explained. So I would like to go to, unless there's any questions, I'll move on to the next issue.

[00:18:32] Speaker 01:
Only a single description?

[00:18:35] Speaker 00:
The single description issue. And I'll start with the substantive arguments first and then go to this procedural argument. So this is, again, another issue that VideoLabs is raising for the first time on appeal. This was not raised below that now this single description limitation somehow requires that there be multiple compatible implementations. But that's not what the claims require. They don't require that. What the claims require is that there be a plurality of different implementations of at least one of the items of digital content.

[00:19:08] Speaker 00:
And then the claim continues that each implementation corresponds to a different set of device capabilities. Now, the claim doesn't preclude that there be multiple compatible implementations per device, but it certainly doesn't require it.

[00:19:27] Speaker 00:
And if the patentee wanted the claims to be limited to requiring that there would be stored – received and stored that there would be multiple compatible implementations for a specific device, then the claims would have been drafted to include that language. But they were not. Even the presenting step where this single description limitation is found, it says – a single description, presenting only a single description of each item of digital content, regardless of the number of implementations of each said items.

[00:20:01] Speaker 00:
It doesn't say regardless of the number of compatible implementations of each said item. So the claims are broader than what Video Labs is now advancing on appeal. And so...

[00:20:14] Speaker 00:
The claims were drafted in the broader sense such that a single compatible implementation for each device would fall within the scope of the claim. And this is consistent with the specification of the 790 patent. The patent says that the product catalog shows, and this is that appendix 81, column 9, starting at line 54, only those products that have at least one implementation compatible with the subscriber's device. So all that's required in the claims is that there just be one compatible implementation per device to meet the claim.

[00:20:50] Speaker 00:
There's no requirement that there be multiple compatible implementations per device. And Video Labs even concedes that if the claims don't require this multiple compatible implementation, then Meta satisfies this And it says by pure happenstance because it would only be in situations where each item of digital content only has a single compatible implementation. But whether it's happenstance or not, that's still anticipation.

[00:21:17] Speaker 00:
And so one of the... Sorry. One of the...

[00:21:25] Speaker 00:
arguments that we heard counsel raise was that this was about this regardless clause, that it would be rendered superfluous if multiple compatible implementations weren't part of the claim. And just as a preliminary matter, this regardless clause being superfluous is being raised for the first time in Video Labs' great brief. This was not raised below, just like multiple compatible implementations was not raised below. But now this is a separate argument that was not raised below.

[00:21:57] Speaker 00:
But putting that aside, The claims only require that the single description of each item of digital content be displayed regardless of the number of implementations. And there's no dispute that meta teaches different implementations. And so if we go back to, you know, I think...

[00:22:18] Speaker 00:
They make this argument, but I think we have to go back to the language in the claims. So if there's no requirement that in the receiving and storing step that there be multiple compatible implementations, and the claim structure is a sequence of events. So you've got your receiving and storing things first, and then it goes to the presenting. And so if at the receiving and storing there's no requirement for multiple compatible implementations, then how could there be a requirement for multiple compatible implementations in this last clause of the claim, the regardless claim?

[00:22:55] Speaker 00:
And so this regardless claim, I think all that really says is that if there's different implementations, like going back to the Microsoft Word example, if there's two different implementations for an item of digital content, then what's going to be presented is just going to be the single description, even though there are two different implementations.

[00:23:17] Speaker 00:
If there are no questions on that, I'll move on to the next, the procedural issue regarding a single description.

[00:23:25] Speaker 00:
So, Bedi Labs claims that Netflix did not present enough analysis on this single description, and that's just not true. If you look at Appendix 128 through 129, that's the petition, that addresses the entire presenting step as a whole. And what Netflix said was that – and it quoted part of Meta, paragraph 117. And what it quoted was that a subscriber could avoid the problem of having to explicitly select a compatible application – When the MOS lists only compatible applications.

[00:24:04] Speaker 00:
So in view of that disclosure, that's talking about exactly what this claim limitation recites. That you're only going to pick the, it's only going to present the compatible applications so that the user doesn't have to go through trying to figure out which version is compatible with its device. And so Netflix cited paragraph 117 of META. And then it also relied on paragraph 77 of Mr. Wexelberger's declaration. and that's at Appendix 227. There, Mr. Wexelberger said, Moss provides the subscriber with only a single subscriber-appropriate version that will work for the subscriber to avoid the problem of having to explicitly select a compatible application.

[00:24:49] Speaker 00:
And so this shows that the petition was not deficient. And I think what's meaningful to know is that if there was a deficiency with the petition, Video Lab should have addressed this in its patent owner preliminary response because they did submit one and they didn't contest this there. And if there was a place to contest the sufficiency of the analysis for this limitation, it would have been then because that could have prevented institution. But they didn't, and because it's not a viable argument, because there is analysis on this specific limitation in the petition.

[00:25:25] Speaker 00:
And so when Netflix, in its reply, then identified in response to Video Labs' arguments, it identified figures 2 and 11-F, that was consistent with the theory it set forth in the petition. And figure 11F is actually an embodiment that is expressly identified in paragraph 117 that was cited and quoted in Netflix's petition. And so what Netflix did was it put forth in its reply analysis regarding 11F and figures 2, which are both the same embodiment.

[00:26:03] Speaker 00:
It's the same embodiment that that's been discussed in its petition.

[00:26:11] Speaker 00:
So if I can just quickly address the obviousness argument.

[00:26:16] Speaker 01:
Well, your time is up, so if you want to give us one sentence.

[00:26:20] Speaker 00:
Sure.

[00:26:22] Speaker 00:
I think the one thing to note for the obviousness argument is that the petition relies on meta, and meta shows that it teaches this limitation. So when the combination of Schlaffer, it just bolsters the analysis and that this limitation was met, and there is substantial evidence to support the board's finding that this limitation was anticipated and is obvious.

[00:26:51] Speaker 01:
Thank you. Thank you, counsel.

[00:26:55] Speaker 01:
Mr. Cardenas Navio.

[00:27:00] Speaker 05:
Thank you. Before I dive back into the merits, I just wanted to spend a minute on talking about the overall inventive concept of the 790 patent because it illuminates some of these disputes that we're having.

[00:27:14] Speaker 05:
The inventive concept of the 790 patent is really this idea of having –

[00:27:28] Speaker 04:
So don't we just look at the claim like the board did and said it's pretty broad and that it shows this.

[00:27:35] Speaker 04:
You want to limit it now based on the overall concept and your particular reading of it. But if you'd wanted the claim to be read more narrowly, you should have asked for claim construction.

[00:27:47] Speaker 05:
I disagree with that, Your Honor, but I'll start there.

[00:27:50] Speaker 04:
I disagree with you should ask for claim construction if you wanted it narrowly. narrower than the plainly meaning. I mean, that's the way claim construction works.

[00:27:57] Speaker 05:
We weren't asking for more narrow than the plain and ordinary meaning. It sure sounds like you are. I will respectfully – we are not. And that's how these disputes ended up getting resolved the way that they did is Videolabs' position from the start was plain and ordinary meaning. Netflix's position was also plain and ordinary meaning. It's not uncommon for that to happen. But then, of course, they had different views of what the plain and ordinary meaning was.

[00:28:21] Speaker 04:
Yeah, and when that happens, you should ask for claim construction. We've said that over and over again in district court cases. And in PTAP cases, that if the parties have a dispute over what the claim means, even if you both think it's plain and ordinary meaning, you need to resolve it through claim construction.

[00:28:38] Speaker 05:
And so, well, I guess I would say in its reply, Netflix could have flagged that as an issue. But either way, the board ultimately resolved the implementation issue certainly as a matter of claim construction. In its analysis, it cites canons of claim construction. It goes through it and analyzes them and ultimately decides the issue by deciding that the scope of the claims is not as Video Labs contends it is and so do not encompass this requirement.

[00:29:08] Speaker 05:
stored association between one item of content and multiple implementations. And I did want to pause for a moment to talk about one thing, which is that Counsel for Netflix framed the dispute as a that the claims don't require how relationships are stored, and we actually agree with that. What we're focused on is that certain relationships are required to be stored.

[00:29:34] Speaker 05:
And I want to turn again to claim two because it's the combination of two requirements. It's the combination of receiving and storing a plurality of different implementations of at least one of the items of digital content with the requirement of the next limitation of storing in association with an in association with each item of digital content, a reference to each implementation. And each, in this case, means at least two or more. And again, our position is not that there's some special or magic way that those relationships need to be stored.

[00:30:09] Speaker 05:
It's just that they need to be stored. And that's not the case in meta. In meta, again, you've got these applications that have a one-to-one relationship, only ever a one-to-one relationship, of one item of content to one implementation, and counsel for Netflix presented a new theory that because different applications can have the same title, that that somehow links that item of content to multiple implementations and multiple applications.

[00:30:40] Speaker 05:
That's not a theory that was presented during the IPR proceeding, and that actually doesn't get them far enough anyway to Because, again, there's no suggestion, there's no teaching in meta that there is, in fact, a linkage that the system recognizes that two applications have the same title and, therefore, their two corresponding URLs are, in fact, linked to the same item of content.

[00:31:06] Speaker 05:
I want to address the presenting limitation here.

[00:31:12] Speaker 05:
Netflix was saying that there is no requirement of storing multiple compatible implementations on the claims, and we agree with that. It's not that multiple compatible implementations can be stored. In some instances, there will be. In other instances, there won't. The point is that there is an act, a second act of filtering as part of that presenting step where if there are multiple compatible implementations, you take them out at that point.

[00:31:39] Speaker 01:
Counsel, as you can see, your time has expired. So we thank both counsel. The case is submitted.