On October 29th, 2026, the New York Times reported that Jensen Huang had a ‘close relationship’ with Alice Huang, an executive U.S. officials had concluded was a Chinese spy. The story had six reporters on the byline, and was sourced to "people briefed on the surveillance."

Whether or not you think Alice Huang was working for the MSS, a rational person might conclude that the "person briefed on the surveillance", who supplied the initial story, was a U.S. government official. After all, if you were not possessing a security clearance and working for U.S. government, you would not be getting "briefed" on U.S. counterintelligence investigations.

As the story gained traction this month, people who think quite differently began to propose other theories. On November 10th - the first time I can find this suggested - an account called " ellington " stated their "personal conspiracy theory" that Anthropic was the New York Times' source (2k likes):

A few hours later GeLun Ding, a Chinese-language account that shows up again below, said the same thing without hedging (1.4k likes):

On November 11th, 6PM Pacific Time, Dream1984Tube, another Chinese-language account announced that he had "found the key clue" to who is really behind the hunt for Jensen Huang, and that he was writing it up:

Twenty minutes later, he says he needs to take another look at Ana Swanson, one of the six Times reporters. At 6:32 PM he brings up Jack Clark (cofounder of Anthropic) for the first time , on the grounds that Clark went to an Aspen Institute dinner with one of the six reporters:

On November 11th at 9:14 PM PT, he posted an AI-generated longform article on X titled " Who Is Really Behind the Hunt for Jensen Huang? " This is the first post, on or off twitter, that accuses Jack Clark of being the source behind the October 29th article, and it's the first one that takes off:

Most of the content is devoted to describing second- and third- degree connections between the people who talked about or expressed indignation at the affair. It contains only the following core points:

At this point in Twitter history, this is the entirety of the concrete evidence offered, that Jack Clark was the source for the October 29th New York Times article. His name has been repeated many times on X throughout the saga. As far as I can tell, no one has ever attempted to explain how Jack Clark would know who Jensen Huang is sleeping with.

Grok's post is then amplified by other mouthbreathers in the tech industry, much more than the previous tweets:

And within the hour, "Hensen Juang", a South Asian twitter account, performs the standard magick whereby inferences are transmuted into observations.

And then, in his most popular tweet, he restates the main insinuation from the article without qualification, and receives thousands of likes:

An hour and a half after posting, he adds a one-word reply consisting of the word "Allegedly", which gets an order of magnitude less circulation than the original tweet.

And a day or two later after that he deletes the original tweet, after about a million people have already seen it:

The rumor continues to spread, in various formats, mostly across other Chinese language accounts (screenshots cropped to fit on the page, so that you can see tweet metrics):

During a podcast , Tripp Mickle discusses how the New York Times first started looking into MegaSpeed last year, for an article published on October 9th, 2025 titled ' Is China Buying Banned Nvidia Chips? ':

Natalie Kitroeff: Okay, so how do we get to this place where one of the most powerful CEOs in the world gets tangled up in such an explosive, high-stakes mess? Like, where does this story begin?

Tripp Mickle: You have to back up to last August [2025] when Ana Swanson and I got a message from a tech industry contact of ours who said, “Hey, you really need to drop everything and look at this company called Megaspeed.” It’s a Singaporean data center firm. I mean, it’s literally a company that buys chips from Nvidia and builds data centers. And there’s nothing about that on the surface that says this is something we should drop what we’re doing and devote time and resources to figuring out what this company is and why it matters.

Context notwithstanding, here is how Philippe Lemoine reports this interaction on X The Everything App, at 6am on November 12th:

Philippe's tweet heavily implies that "the story" that "didn't originate with the intelligence community" is the one that kickstarted Jensen's firing. A casual reader would incorrectly assume that a "tech industry contact" is who told the New York Times about Jensen's affair, which is naturally what Twitter users do.

The next day, Chief Nerd, an account with nearly a million followers, cites the same clip but this time claims that the "Jensen Huang story" was the one tipped off by the "tech industry contact:"

And about this time is when Chinese language accounts tweet the same link, transmuting "a tech industry contact" into "Anthropic", in order to get an order of magnitude more uptake:

This post, by "Dragonhau66", is what is finally screenshotted and converted back into English by other Chinese-speaking users. During the translation, "Anthropic" is transmuted further into specifically "Jack Clark, co founder of Anthropic":

And about then is when the "story" really takes off:

It's probably worth noting that back in the land of legacy media, Marco Rubio was already named as the author of the memo warning Trump about the relationship between Huang and a suspected Chinese spy . According to the Washington Post:

Rubio used the CIA report to write his own, recasting the agency’s reporting in his own more assertive terms, the person said, speaking on the condition of anonymity to describe confidential intelligence.

The Post article describes the memo's authorship as part of a power struggle within the executive branch, over whether Nvidia should be allowed to sell chips to China. I don't think Rubio leaked the memo, but if people feel like being conspiratorial, it seems much more plausible that he circulated the memo knowing there was a chance it would be leaked.

It is now Sunday and the story is still ongoing; I saw more and more people casually mentioning the "Anthropic coup" while doing research to write this post. I suspect I will continue to see references to Jack Clark's treachery for months to come. Doing this trace took way more time and energy for me than it did for these individuals to write their tweets in the first place, and my trace will not spread nearly as far. It certainly wouldn't change anything if I took the time to reply to every one of these examples with my lowbie account.

I don't really care about the particular topic, or about Jack Clark. I only take the time to discuss this pattern as a way of showcasing how information travels on Twitter in the modal case, and to gesture at the hundreds of other well-circulated bits of false information I've observed, which no one ever bothers to "debunk" in some visible way. If you use Twitter to keep up with AI, especially if you tell yourself that you're "protecting" against the malicious influence of traditional media, you too will fall for slightly more plausible examples of this pattern. No one is going to take the time to debunk them for you.

Even if the information on twitter is often false, I'd still vastly prefer having one place on the internet where we can have a lot of false information, than having traditional media be the entire arbiter of truth. Without twitter, we'd be stuck in the lies and propaganda of traditional media entirely. In this case. I believe it's directionally correct to model what happened as some type of power struggle in the government between factions that were okay with letting jensen do what he wanted to do, and factions that weren't. And that jensen got ousted as he basically lost this power struggle. This is just normal politics. So although the specifics of what twitter is saying in this case are probably incorrect, we don't have to accept the official story . I believe it's already not clear if she was really a spy for example? Wasn't that the official story? Or is it an 'alleged' spy? Whatever either way.

I think the main reason to worry about “traditional” media is the possibility of a biased and censored central authority. Yet, I think state level actors can still exert huge influence on decentralized media - pushing stories they want to see with bots or sock puppets, under the cover of “free journalism.” Let alone the power of who sets the algorithm.

Accusations against Anthropic (and Clark) made by users on X are speculation, not confirmed facts. I hope most people are aware of that. Do we have any further evidence? Well, political donations could be considered tips pointing in that direction.

The problem is that there is no evidence of espionage, yet the case is receiving so much media attention that it’s no wonder people are starting to wonder what exactly is going on. Robert Wright—a Pulitzer Prize winner and advocate for a global agreement to halt the development of the most powerful AI models—wrote on his Substack:

China hawks in the national security establishment are pushing their agenda fiercely and effectively, and that doesn’t bode well for the US-China cooperation that I think will be critical to human welfare as the AI revolution unfolds.

There is graph (but probably AI generated, idk), with plenty of connetions between Anthropic and american politicians that want to take down Huang. For example:

Chuck Schumer, the Senate Minority Leader, had advisor ( Mike Kuiken) that currently sits on Anthropic’s advisory council. Senate is set to summon Huang for questioning. But there are other connections. But it seems to me that people generally have the good sense not to take the alleged roles of Rubio, Anthropic, and Clark as 100% confirmed facts, right?

Thank you for your sacrifice in tracking this down.

I only take the time to discuss this pattern as a way of showcasing how information travels on Twitter in the modal case

In the same genre: my attempt to map out the early-2025 freakout about some alleged insane breakthrough at OpenAI.

I suppose one annoying counter-example to this pattern is September 2026 rumors about a frontier lab resolving a Millennium Problem, which were also warped, but ultimately did originate from something true (the Levent/Tristan collaboration leaking). That unfortunately means that "Twitter noise is always bullshit" is not quite the right model to have... But you should probably internalize, at the gut level, that it ought to be the default interpretation (instead of "there's no smoke without a fire"), and you should only dig into it if you really have nothing better to do.