Mr. Zuckerberg, please raise your right hand. Do you swear or affirm that the testimony you will give today will be the truth, the whole truth, and nothing but the truth?
Good day, Mr. Zuckerberg. The law must apply equally to everyone, right?
Regardless of how wealthy you are, the law must apply to you the same way.
Regardless of how smart you are, the law must apply the same way.
And it shouldn't matter how big or powerful a company like Meta is, the law must apply to it the same way.
I believe that's correct. The law must apply equally to everyone.
Okay. Meta Platforms is one of the few companies in world history to achieve a valuation of $1.5 trillion. Is that correct?
I believe that's correct.
And Meta announced a goal of continuing to grow and attempting to increase its value by 500 times to reach $9 trillion in the next five years by 2031. Is that correct?
I don't believe that's correct. I'm not sure what you're referring to.
Okay. Can I have the first binder, please? And what did we decide to mark this as?
Mr. Zuckerberg, we've handed you Plaintiff's Exhibit 572. Does that refresh your memory that Meta has a $9 trillion goal by 2031?
I'm not sure what you mean. You asked whether we had announced a goal to increase our valuation by 500 times, and I don't believe that's correct.
And did I... I misspoke.
Okay. Here it is. Do you see that shaded part there? From there we got that.
Okay. Yes, I mean, I don't think that was an announcement we made.
Okay. Mr. Zuckerberg, do you have a $9 trillion goal by 2031 for Meta?
I'm not sure what you're referring to. We have an executive incentive program where they earn certain amounts of compensation at reaching different milestones. I believe one of those milestones roughly corresponds to what you're saying, but there are other milestones too. This wasn't a program we announced as a company goal. It's just a program we have as part of executive compensation for some people in the company.
Okay. So executives will get more money if they reach a $9 trillion valuation by 2031. Is that a fair way to say it?
That's part of the program, but I think it's also fair to note that people across the company and our shareholders generally benefit when the company is successful.
Okay. You're the CEO of Meta Platforms.
And you were the CEO of your previous company, Facebook. Is that correct?
As CEO of Facebook, did you found the company?
You were responsible for what happened there. Is that correct?
And you had ultimate authority at Facebook. Is that correct?
Well, I mean, I have the authority that comes with being CEO.
Right. And during your time as CEO of Facebook, you always held a majority of the shares in Facebook, didn't you, sir?
I don't believe I own a majority of the shares.
Yes, I have a level of shares that gives me a majority of the voting power, which I think is what you may be referring to.
That's what I'm referring to.
I don't own a majority of the shares.
And because you own a majority of the votes, you could appoint or fire the board of directors, couldn't you?
I suppose that because I own a majority of the votes, I can win most shareholder votes.
Okay. And because of your control of the company, you had the advantage of not having to convince the board not to fire you, is that correct?
I believe that's generally correct. I've done some interviews on podcasts where I've made this point, although I think it's somewhat simplistic. I mean, of course, the board has the ability to fire me if they think I'm not the right person to run the company or if I'm not doing a good job. But generally, I mean, I have voting control of the company.
And if the board wanted to fire you, as the person who owns a majority of the votes in the company, could you elect a new board and reappoint yourself, couldn't you?
In theory, over a period of time. I think it would be chaotic.
Okay. Isn't that exactly what you testified to in the California case on February 18, 2026?
Do you want to show me what I said?
Sir, you were asked a question. You said: 'Because I control my company, I have the advantage of not having to convince the board not to fire me.' That was an interview you did with Joe Rogan. And you were asked: 'Do you stand by that?' And your answer was: 'I believe that's roughly correct, because if the board wanted to fire me, I'm the person who owns a majority of the votes in the company, and I can elect a new board and reappoint myself.' Isn't that what you testified to in February of this year, sir?
Yes, and that's also what I just said.
Okay. So practically speaking, you can't be fired.
No, I mean, I think I've explained this before. The board can fire me. I think your question was right that if I wanted to, after being fired, I could theoretically elect a different board because I'm the controlling shareholder of the company, but that doesn't mean they can't fire me or that it wouldn't be a very complicated and chaotic process.
Okay. You could stop things at the company that you didn't agree with, couldn't you? Did you have that authority?
As CEO, I think most CEOs have that authority. There's always the question of what you're exactly aware of happening inside the company, but yes, in theory I suppose most CEOs can prevent things from happening inside the company that they believe are the wrong thing to do.
So my question was simple. You could stop things that you didn't agree with, sir. Is that correct?
The promises you made as CEO of Facebook were promises on behalf of Facebook, weren't they?
I believe that's correct.
And the promises you're making now as CEO of Meta Platforms are promises on behalf of Meta, aren't they?
If I make a commitment in my professional capacity, it's a commitment I make on behalf of the company, yes.
One of the things you promised as CEO of Facebook was to apply the rules that Facebook created equally to everyone. Isn't that correct, sir?
That sounds like something I might say, but can you show me what you're referring to?
Can I have Tab 5, please? These are Exhibits 204. Mr. Zuckerberg, you testified before the United States Senate Commerce, Science, and Transportation Committee on November 28, 2020. Is that correct?
This document says October 28.
Sorry, I got the wrong date then. October 28, 2020. Is that correct, sir?
That's what the document says.
Okay. And were you under oath when you testified before the Senate Commerce, Science, and Transportation Committee?
I suppose so. I don't remember this well, but I can read it if you want me to answer questions about it.
And I'll ask you about what you said previously to this Senate committee. You were asked, Mr. Zuckerberg, can you commit that if the president went on Facebook and encouraged violence after the election results, you would ensure that your company's algorithm would not amplify harmful content, and that you would take those posts down immediately? And your answer, sir, on the next highlighted page, was 'Yes, Senator.' 'Inciting violence is against our policies, and there are no exceptions to that, including politicians.' Is that what you testified to, sir?
That's what's written here. So I have no reason to believe that's not correct.
And Senator Markey asked you, 'Are there any exceptions?' Did you say that? And you said, 'There are no exceptions,' didn't you, sir?
I'm not sure if that's a transcription error. I believe this... I agree with what's written here. I'm not sure that I would have said 'there are no exceptions.' Maybe I would have said 'there aren't any exceptions,' but...
Okay, you're just quibbling with grammar.
Okay, thank you. Then Tab 6, please, Gordon, which is the Plaintiff's Exhibit... the State's Exhibit 203. This is State's Exhibit 203, and it's very small because that's how it was printed, but we enlarged it in the back. There are two parts I want to ask you about. Do you recognize that as a picture of your Facebook page?
And when you wanted to get things out to people, you posted things on Facebook yourself, didn't you, sir?
Yes, that's one of the ways I communicate.
And when you post on Facebook, do you write the posts yourself?
Are those posts shown to your communications team?
Depending on the content, I show them to different teams internally.
Okay. This is a post from June 26, 2020. And if you look at the enlarged part, one of the things you said in this Facebook post is that even if a politician or government official says it, if we decide that content could lead to violence or disenfranchise people, we will remove that content. Similarly, there are no exceptions for politicians in any of the policies announced here today. Isn't that what you said, sir?
Are you asking about the part I highlighted here?
And your question was just whether this is the post?
My one question is whether this is what you said in your post.
Yes, I was just asking whether your question was whether you were asking me if you were reading part of the post correctly?
Okay. And if I could have Exhibit 202, please. I only have seven. Mr. Zuckerberg, do you recognize Plaintiff's Exhibit 202 as another one of your posts?
Mr. Zuckerberg, didn't you say in that post that our current policy is that if content is actually inciting violence, the right thing to do is to remove that content, not to allow people to continue seeing it behind a warning label? There is no exception to this policy for politicians or for its newsworthiness.
You're reading what's in the post correctly.
And if we could have Tab 8, please. I want to ask you about Facebook Newsroom posts, because that was another way Facebook communicated with its users about its policies. Is that correct?
Yes, the Newsroom is like our company blog.
Okay. I want to ask you about one of your posts from there. This is Exhibit... thank you, counsel... 201. And I want to ask you about what's on the first page.
So I want to ask you about two parts of that first paragraph.
Okay. And I want to ask you about the first sentence first, which is highlighted in yellow.
Okay. Are there other parts of this you want to ask about? Okay. Okay. So, just this. Okay.
It's under the heading 'Cross-check,' and the first sentence says: 'We want to be clear that we remove content from Facebook regardless of who posts it when it violates our standards. There is no special protection for any group, whether from the right or the left.' Isn't that what it says, sir?
Yes, you read that correctly.
And that's what Facebook was putting forward as its policy on that date. Is that correct?
I believe that's a fair interpretation.
Okay. And the date on that was July 17, 2018.
And the second part of that talks about something called 'cross-check.' And I'll talk to you more about that in a moment, but 'cross-check' often in Facebook records, instead of the word 'cross,' it has an X and the word 'check' after it. Is that correct?
I believe I've seen it referred to that way.
Okay. And this part says what's highlighted in blue there: 'Cross-check, the system described in the reports simply means that some content from certain pages or profiles is subject to a second layer of review to ensure we applied our policies correctly.' Did you read that correctly, Mr. Zuckerberg?
Okay. Thank you. The people who paid to advertise on Facebook didn't like that their content was checked or removed, did they?
I'm not sure what you're referring to.
Okay. Do people like having their content reported or removed?
I believe that's a very standard part of how the service works.
Yes, sir. Do people like having their content reported or removed?
I'm not sure that anyone likes being told their content is unacceptable.
Okay. Among the powerful people who complained about their content being removed, reported, or downranked, there were politicians. Among the powerful people who complained to Facebook about their content being removed, reported, or downranked, there were politicians.
I believe that's correct. I mean, a very wide range of people have complained about things they believe limit free expression, and that includes politicians, but I don't think it's limited to them, as you know.
To review the list of other people who also complained. Politicians who hadn't held office yet, but who were running for office, complained that their content was...
Yes, among many other people.
Political commentators who were commenting from both the right and the left complained about their content being reported, didn't they?
And celebrities who had many followers on Facebook complained about their content being reported or downranked, didn't they?
And despite your promise to treat politicians like regular users, Facebook created what's known as a whitelist, which protected politicians from fact-checking, didn't it?
I'm not sure what you're referring to.
Okay. You're not aware of the whitelist at Facebook that protected politicians from fact-checking?
I'm not aware of what you're referring to as a whitelist. I believe the terms related to the misinformation and fact-checking policy specifically directed fact-checkers on what content was eligible to focus on, and we were trying to focus on bad and harmful hoaxes, not analyzing, as you know, the words in political speech. So I believe that was the explicit guidance we gave fact-checkers because we wanted to focus on things that were large-scale hoaxes.
So you told them not to fact-check politicians?
Well, we told them the areas we believed they should focus on, and yes, I believe we told them explicitly that the goal of the fact-checking program was not to focus primarily on analyzing the words of politicians, but on preventing the spread of misinformation at scale on the platform.
You actually created a list in 2020 of more than 4,300 individuals and entities in the United States who were not subject to fact-checking, didn't you?
Sorry, when you say 'you,' I assume you don't mean me personally, but the company. In fact, I'm not entirely sure.
Okay. Can we have, let's see, Tab 12 and Tab 10, please. Thank you. So, Mr. Zuckerberg, we submitted these... I'm sorry, Hannah. This is Plaintiff's Exhibit 598. We submitted questions to Meta during discovery, and we asked your company for answers about both the whitelist and the cross-check program. And these are the responses your company provided. If you would turn, sir, to page 10, we asked about Facebook's political whitelist, and the answer was that as of November 3, 2020, Facebook maintained a whitelist of politicians that included more than 4,300 individuals and entities in the United States. Do you see that, sir?
Yes, I see it. Do you want me to read the rest of this document?
Not yet. No, sir. This is the first part I'll ask you about. And now, if we could have... yes, Ted. Yes. This is Exhibit B. Attached to it, sir, was the list of the 4,300 individuals, which was Exhibit B in the discovery responses. This is Plaintiff's Exhibit 206, Hannah. And sir, do you recognize this as a political whitelist maintained by Meta on November 3, 2020, which included politicians who were not subject to fact-checking?
I'm not sure I've seen the list before, but it seems logical to me that if our policy was to have fact-checkers focus on large-scale hoaxes and not on political speech, we would have to implement that in practice by creating some kind of list that directs fact-checkers on the types of material that should and shouldn't be reviewed.
Okay. So you don't deny that this was the list Facebook had in 2020.
Yes, I mean, I don't deny that, but I'm also saying I don't believe I've seen this before. So I can't confirm that either.
You weren't one of the fact-checkers who used this list to know who was exempt from fact-checking.
No, I wasn't a fact-checker.
Okay. And there were more people and entities on this political whitelist who were outside the United States. Isn't that correct?
I haven't reviewed the list, so I can't speak to that precisely, but you asked about groups like politicians. My understanding is... are you asking specifically about this misinformation fact-checking issue, or...
Let me, let me rephrase my question for you, Mr. Zuckerberg. In your response to discovery, what Meta confirmed was that these are the 4,300 individuals and entities within the United States. My question is simple. Were there other lists of politicians outside the United States? Isn't that correct, sir?
What are you referring to?
I'm referring to this document, number 598. Page 10 of it. That this document (B) represents only political figures in the United States.
So my question is just whether there were other lists of politicians in other countries. Isn't that correct, sir? Who were also exempt from fact-checking.
I'm not sure I'm the right person to answer this question. I mean, I believe you can find this information elsewhere, and I don't know what lists we were maintaining.
Okay. The people on this list, weren't they treated differently from regular users?
No, I don't believe that's correct.
Okay. Over time, the list of politicians on the whitelist grew longer, didn't it?
I've already testified several times that I don't really have knowledge of what was on that list. So I certainly don't know how it changed over time.
Okay. If you would turn to page 6 of that document you have now, sir. Those are your responses to the interrogatories. We asked for the whitelist as of 2025, and the response Meta gave was... Document A, okay, two responses. Facebook maintained a whitelist of politicians on August 29, 2025, containing more than 5,500 individuals and entities in the United States. Do you see that, sir?
I see it, but I believe I need to read this document if you're going to ask me more questions about it.
Okay, if you'd like to take a break, these are the responses your company provided.
I have no reason to believe that's not correct, but if you want to ask me questions about something I'm not aware of, I need to have time to read it.
Okay. Okay, let me ask you about Document A on page 6, which is also highlighted. Document A, if I could have that as well, Gordon. Is this the list of individuals and entities in the United States that appeared on the political whitelist closest in time to August 29, 2025, which bears Plaintiff's Exhibit number 205? And my question to you, Mr. Zuckerberg, do you object to your company's responses to this discovery, and is this the whitelist that grew by 2025, the political whitelist?
I have no reason to believe any of this is inaccurate, but you're asking me questions about things I'm not personally aware of the details of. So if you want me to answer the questions, you'll need to give me time to read the documents.
Okay. Let me ask you another question. The whitelist program wasn't the only system Facebook and Meta created that treated people differently, was it?
I'm not sure what you're referring to.
Certainly. The other was the Cross-check program, or the XCheck program, as it's referred to in some of your documents. Is that correct?
I mean, I disagree with the characterization in the question. This isn't treating people differently if we say we have a policy about general media misinformation and not about analyzing political speech.
Okay. No, no, no. I'm just saying this is the same program, the same program that was referred to in the Newsroom post we discussed earlier. Is that correct?
Are you asking whether what you're now calling 'Cross-check' is the same as what's referred to as 'Cross-check' here?
Okay. Let's talk about who was on the Cross-check list instead of the politicians list. Were there business partners making money for Facebook on the Cross-check list?
I'm not familiar with the entire list. So I certainly wouldn't be surprised if there were people on the list who were business partners, but I can't name any of them from memory.
Okay. Were there celebrities on that Cross-check list?
Were there podcast hosts and other influencers who drove engagement on Facebook on that list, the Cross-check list?
I believe so. And more broadly, I believe it focused on sensitive accounts that we were worried about making mistakes in reviewing, so we wanted to make sure they were reviewed twice. So it also included people like human rights activists and journalists and people who might have adversaries trying to report their content or take it down, and we just wanted to make sure we were getting the right level of review. But, to be clear about this, because I think you're hinting at something here. The rules and policies are exactly the same for people on the Cross-check list and for people not on it. The purpose of these sensitive accounts was to ensure we reviewed the content accurately, to be very careful not to make a mistake by leaving up something that should have been taken down, or taking down something that shouldn't have been.
Okay. Do you agree with your head of civil integrity, Sumedh Chakrabarti, that anyone in management, including you and Ms. Sandberg, could request adding people to that list? The Cross-check list?
What are you referring to?
I'm referring to his testimony where he said there was a group of 'Friends of Mark' and 'Friends of Sheryl' added to this list at your request. Did that happen?
I'm not aware of that, but if you want to show me testimony or documents about this, I'd be happy to read them to see if they refresh my memory, but for now, I don't recall anything like that.
The Cross-check list included people who contacted you and asked to be added to the list. Is that correct?
One of them was Joe Rogan, the right-wing podcaster.
Again, I don't... that's possible. I don't recall that.
Mr. Rogan contacted you personally in 2023, didn't he?
And complained about Facebook removing his content?
I believe we made mistakes in handling his content and took down things we shouldn't have. I don't recall an interaction about his request to be added to any program. It's possible that happened, but I don't recall it.
After that, you issued instructions to the people working for you...
Sorry, counsel, could you say that again without...
Certainly, certainly, certainly. After that, you issued instructions to the people working for you that he should be added to the Cross-check program.
I don't recall that, but it's possible that if we made multiple mistakes on his account, I might ask them to take steps to try to reduce making more mistakes.
This is Trial Exhibit 261. And if you could turn to the second-to-last page, Mr. Zuckerberg, this is where we'll start. This is an email from you, Mr. Zuckerberg, is that correct?
Okay. What's your question?
My question was, Mr. Rogan contacted you directly complaining about being banned and his posts being deleted. Is that correct, sir?
I believe what you're saying is roughly correct, although, to be clear, I'm not sure the account he referred to was his personal account. It might be an account run by someone else, I'm not sure.
He said Joe Rogan emailed me to tell me that posts on the companion 'Joe Rogan Experience' page on Instagram were being banned repeatedly.
Yes, and I'm saying I'm not sure whether he runs the companion 'Joe Rogan Experience' page. I believe this might be about someone else posting on his behalf, but I believe your point is roughly correct. I'm just trying to be precise.
Okay. And you also asked and pointed to whether we have a list of accounts that we double-review before taking any action on them. Can we add Joe's pages to that list to make sure we reduce false positives on his accounts as much as possible? Do you see that, sir?
Let me ask you about the email. Was email one of the ways people at Facebook communicated? Is that correct?
And it works like any email anywhere, the person with the email address is usually the one who writes the message at the time it's written. Is that correct?
And the way you got work done at Facebook was through these emails. Is that correct?
Yes, email is one of our communication channels.
And you didn't delete those emails. You kept them as part of your business records. Is that correct?
And this record is one of the records that was preserved in the same way. Is that correct?
I assume so. And that's why it's in your possession.
And if you could turn to the first page, on September 22, 2023, a person named Mathias Storm, am I pronouncing his name correctly?
Okay. He sends you an email saying: 'We're pleased to inform you that Joe's relationship manager has the complete list of accounts managed by Joe's team, and the relevant XCheck tags have been added to those accounts to ensure a second review before enforcement.' Is that correct?
I believe you're summarizing that correctly.
Okay, thank you. Let's talk about what the Cross-check program did. The Cross-check program prevented the immediate removal of content that violated your rules. Is that correct?
Okay, my understanding is that the purpose of this program is to make sure sensitive content is reviewed twice before we take any action in either direction.
Yes, sir. So it prevented the immediate removal of content, didn't it?
I believe that's an inaccurate description of the program's purpose and how it works, but it required a higher level of review of the content to ensure the right decisions were made about the content.
And that means it wasn't removed immediately.
It was removed as soon as that process was completed.
Okay, but not as soon as it was flagged by your AI program. Is that correct?
I believe that's correct, although there's a lot of content that's flagged by AI for a person to review. So I don't agree with the implication in your statement that this is somehow unusual.
Even if your AI program classified a post as hate speech or containing violent or nude scenes, or inciting violence, or containing harmful information, if the person is on the CrossCheck list, it won't be removed immediately, will it?
I mean, we have a lot of content that AI flags for people and employees review because it's important. And in these cases, if you're dealing with human rights activists or journalists, people who have adversaries trying to report their content and claim there's a problem with it when there often isn't, it seems appropriate and important to make sure there's an additional review step to ensure we're handling those matters correctly.
So again, my question is just, if you're on the CrossCheck list, even if you're in one of those categories, like hate speech or violence or nudity or inciting violence, if you're on the CrossCheck list, the content won't be removed immediately, even if it's flagged by AI, will it, sir?
I mean, I believe I've explained what the process is, but your question implies that any other content would be removed immediately, and I believe there are various ways we handle this, including AI often escalating to employees, even if the people aren't on the CrossCheck list. So I believe you're hinting at something I believe is incorrect in your question, but I've already explained the process.
Let me rephrase the question for you then. Where an ordinary person gets a violation or is immediately put in 'Facebook jail,' you've heard that term, haven't you, sir?
People refer to it and use it to refer to different things, so you might want to be more precise, but I've heard the term.
Okay, while an ordinary person gets a violation, the person on the CrossCheck list...
that is on the cross-check review list for additional review opportunities. Is that correct?
Again, I don't agree with the way you're describing this. I mean, the purpose of the program is to make sure we don't make mistakes in content review.
The more important the person or organization on the cross-check list was to Facebook, the more levels of review they got. Is that correct?
I'm not aware of anything like that. I thought we had a standard process for this.
Okay. You weren't aware that some of them were contacted and given the opportunity to remove content so they wouldn't get a violation on their accounts that could lead to losing their privileges?
I'm not sure what you're referring to.
Okay. Haven't you heard the term "firefighting" for a member of the cross-check list, which means they got three levels of review before content was removed or its spread reduced? The term "firefighting," meaning people on the cross-check list got three levels of review before content was removed or its spread reduced?
I don't think I've heard that term until now, when you said it.
Some people's content wouldn't be removed even after three levels of review unless someone in management agreed to remove it. Were there people like that on your list?
Again, I'm not sure I'm aware of all the details of how the process related to that works.
Well, the problem with treating some people differently in the cross-check program is that the review levels — whether it was one, or three, or three plus a manager or someone in management — took time to conduct that review, didn't they?
Sorry, I'm not sure I understand your question. Are you saying what's the problem with that, or are you just saying that conducting the review takes time?
Yes, sir. The problem was that if you had — whether it was one level or three levels or three levels plus management review —
conducting that review would take time before a decision was made about whether the content should be removed.
I mean, making content review decisions takes time across the entire company.
And a video that violates policy and contains hate speech or violent content or harmful misinformation can actually spread like wildfire within hours of being posted, can't it?
That can happen, but we also — from an operational standpoint — pay more attention to fast-spreading content that gets wide distribution compared to content that not many people see. So I think what you're implying is that important content might be delayed more. I don't think that actually happens that way in practice.
Okay. Well, let's — we'll talk about that more in a bit. Even though your AI tools had the ability to detect violating content immediately and remove it, they didn't do that for people included in the cross-check program, did they?
I mean, I feel like I've already answered this several times, but we have AI tools that try to identify content across the company, and in some cases they have the ability to remove content, and in many other cases they flag it for human review. The program we're talking about here, I think, is an example of cases where content is flagged for human review, but there are many examples where the AI system may not be confident enough or where we may need additional review before taking any action.
By 2022, the additional reviews had piled up and were taking about five days on average, isn't that correct, sir?
I'm not sure what you're referring to. We've had different backlogs at different times. So it's possible there was a backlog in this process. But I also think we've had backlogs in other parts of the company, and we generally work to reduce that as much as possible.
Five days is a lifetime on the internet, isn't it?
I'm not — I'm not sure what you mean. I mean, this...
Okay. Fast-spreading content will have already had its impact and been widely distributed, and the viewer will have moved on to something else after five days, won't they?
Well, yes, certainly the lifespan of content can be shorter than five days if that's when most of it is viewed. But again, as I said before, when something spreads very quickly, we're more likely to prioritize seeing and reviewing it faster. And most of the time, when not many people are viewing certain content, we let it rank lower than things that many people are viewing.
One of the groups on your cross-check list was the right-wing news outlet Breitbart, wasn't it?
I don't know that as a fact, but as a prominent outlet, it seems likely they could have been.
That site was one of Facebook's managed partners.
Again, I actually don't know the full list of all our managed partners, but as a prominent news site at the time, I think it's likely they were.
In July 2020, during the pandemic, Breitbart created a livestream video containing false information about COVID-19 in violation of Facebook's policies on misinformation and harm, didn't they?
I think you'll need to show me the specific content. I'm not aware of that.
I'll show it in a moment. This is Plaintiff's trial exhibit 258. Mr. Zuckerberg, this is one of the emails preserved within Facebook. Is that correct, sir?
I believe that's correct.
And it discusses Breitbart's COVID misinformation and the harm it caused. Do you see that, sir, in the first paragraph there?
Let me read it. Okay. What's your question?
My question was that the content regarding Breitbart's COVID misinformation and the resulting harm was immediately flagged by your AI system. Is that correct?
This summary, which I believe I'm reading for the first time and I'm not part of this email thread, indicates that it was flagged by the system, and it appears the cross-check system requested more review, then it was escalated to someone else, and ultimately enforcement action was taken. But I think the point of this is that we're constantly trying to improve our enforcement procedures. So I think I'm reading this email as trying to describe a way we could handle enforcement faster, and the team is working on doing that, which is an important part of our overall process.
Okay. By the time the cross-check review was completed, the livestream had already gotten 40 million views, hadn't it, sir?
It says here that — and I think the abbreviation VPV means it appeared in the News Feed. That doesn't necessarily mean those people watched the entire livestream, which I think is probably your objection, but 40 million people — it could have appeared in front of at least 40 million people who scrolled past it in the Feed.
Okay, thank you. I want to talk about the number of people exempted or included on your cross-check program list, and if we can go back to the deposition answers, which is this document 598. We asked about the number of people on your cross-check list as of November 3, 2020, and your answer was that it contained 13,500 individuals and entities and groups and pages located in the United States. Do you see that, sir?
I see what you've highlighted, but I'll read around it as well.
Okay, I can bring it back. D2.
We're handing you what's been marked as Plaintiff's trial exhibit 208. This is exhibit D2, Facebook's responses to interrogatories, and it's a list of 13,500 individuals who were on the cross-check list as of November 3, 2020. Do you see that, sir?
Okay. Then that group of 13,500 people in the United States was treated differently from the rest of your users, the regular users on Facebook, wasn't it, sir?
The content review process was handled with more sensitivity, but the rules applied were the same.
Oh, I see. Okay. So, can you turn to page 15 of the interrogatory responses again, sir? We asked you about the number of people who were on your cross-check list after five years, on August 29, 2025. That's on page 15, sir. Have you gotten there?
Okay, I've gotten there. What are you asking me?
And after five years, the cross-check list in the United States had expanded to include 799,000 individuals and entities and groups and pages located in the United States. By 2025, the cross-check list had expanded to include 799,000 people who received an extra level of review that regular users didn't get, isn't that right?
I mean, I can just read from this document. Facebook maintained a cross-check list of more than 799,000 individuals and entities and groups and pages located in the United States, consisting of individuals and entities and groups and pages that received an additional level of review after content was identified.
Okay, and you don't dispute that, do you, sir?
No, I assume this was reviewed with extreme care and great precision.
And on the next page, 16, you provided Exhibit C, which was the list of individuals and groups and entities and pages. Exhibit C is what Facebook provided as the list of individuals and entities and groups and pages on the cross-check list as of August 29, 2025. Is that correct, sir?
That's what the text says.
And if we could get Exhibit C, please. That's correct.
Do you mind clearing some space so I can put this down?
We can push that maybe here. Yes, okay.
I'm done with this. Okay.
Okay, Mr. Zuckerberg. This is Exhibit C that Facebook Meta provided as a cross-check list as of August 2025. This is trial exhibit number — let me see — exhibit number 207. Trial exhibit 207. And on that cross-check list from 2025, if you would turn to — okay, first of all, there are about 164 people on each page of those pages. There are about 5,000 pages of names on this cross-check list that was provided. And this number, 799,000 people, could fill the largest 10 stadiums in the United States. This number of people could fill the largest 10 stadiums in the United States. Is that correct, sir?
Objection, relevance.
Nearly 800,000 people.
I don't know how many people the largest 10 stadiums hold, but I haven't seen the document before, so I can read it if you want, but I think that will take longer than we have today. But, I mean, our team is trying to do comprehensive work and review content with sensitivity, and I assume that's what we were trying to do here: make sure that if someone is trying to report something inappropriately, we have the right level of review to reduce the mistakes we make.
Okay, they got a different review than regular Facebook users, didn't they, sir?
We have different review processes for all kinds of things across the company based on the level of sensitivity and the probability of making mistakes. We're trying to reduce the number of mistakes we make, whether by removing content that shouldn't be removed or leaving up content that shouldn't be left up, but it's a hard problem to solve correctly. I'm very proud of the systems we've put in place. We have more than 3 billion users on our platforms. So while you could say that a few thousand or tens of thousands of people is a large number in absolute terms, it's a very small percentage of the overall community. And when there are sensitive matters, I think having a process that allows us to review and handle them with more sensitivity is appropriate.
Okay, let's talk about some of the people on that list who got extra review. If you could turn to page 119 of that list, they're all in numerical order, sir. So, 119 —
On page 119, Alex Jones was one of the people who received extra review. Alex Jones was one of the people on your cross-check list who received an extra level of review that regular people didn't get. Page 119, sir.
What is this? Did you print the same thing over and over to make it look like more pages?
No, sir, I don't think so. Okay, let me ask a question without you looking for it, and let's see if you remember, Mr. Zuckerberg. Was Alex Jones one of the people on your cross-check list?
Well, I don't remember who was on the cross-check list. So I'd have to look into this.
Okay. What about Andrew Tate? Do you remember whether he was on your cross-check list?
Well, again, I don't remember specific people on those lists.
InfoWars? InfoWars, were they on your cross-check list?
Well, I'm not sure. Again, I didn't run the list, so I'm not sure I'm the right person to ask, but you have all the information, so I think it's out there.
Do you know whether the Oath Keepers were part of your cross-check list, sir?
Oh, same answer. I don't know the details.
Mr. Zuckerberg, in October 2021, the Wall Street Journal published an investigative series called "The Facebook Files." And one of its stories revealed and criticized your cross-check program. Do you remember that, sir?
Well, I think I remember something like that, but not in detail. Do you have the document?
I don't have that document, no, sir. Well, by that time, 2021, Facebook had something called an Oversight Board, didn't it?
I don't remember the exact date, but we did form an Oversight Board.
Oh, it was a group of world and American leaders that you hired to review Facebook's decisions?
Well, I'm not sure that we hired them.
I think the idea was that this is an independent entity. So I think the Oversight Board hired its own members. Well, but the purpose of that was to create a level of independent governance as a final step of appeal for people who disagreed with our content decisions.
Well, did you pick the people appointed to the Oversight Board? Facebook or Meta platforms?
I think we picked the chairs. Then I think that group picked the rest. There may have been some consultation, but I think there was a degree of independence there.
