“It’s Blood Money, (But) I Like Money”: Viral Anthropic Quitter Jacob Coxon Refuses To Donate His AI Stocks

Jacob Coxon walked away from Anthropic warning that the AI industry was gambling with humanity’s future, but he isn’t walking away from the money he made helping build it.

Speaking with Stanford professor Rob Reich at an event hosted by Stanford HAI and the AI Governance Lab on November 9, the former pretraining researcher admitted that he still holds equity in OpenAI, agreed that it could be described as “blood money,” and gave a remarkably candid explanation for why he’s keeping it anyway.

“I like money, bro,” Coxon said.

jacob coxon anthropic

“It’s Nice Money. Like, I Like It.”

The exchange began when Coxon was asked about the narrative that AI safety warnings from industry insiders are really a form of regulatory capture, an attempt by incumbents to shape rules in their own favour. That criticism has been levelled at frontier labs repeatedly in recent months, and Coxon himself has been at the centre of a debate around whether his viral resignation post was entirely organic.

Coxon conceded that his financial stake in OpenAI made that argument harder to rebut.

“Um, I think the regulatory capture narrative is a little bit, uh, a little bit confused,” he said. “I agree it would probably be easier for me to argue with that if I like, you know, ceded rights to my OpenAI stock, but it’s like, it’s nice money. Like, I like it.”

Reich pressed him. “But you could take the moral high ground,” he said.

“Yeah, I could. I could,” Coxon replied.

“But you’re not.”

“I’m not, no. Why not? I like money, bro,” Coxon said. “I agree it’s blood money.”

“Do You Prefer Dying To Liking Money?”

The interviewer then went for the jugular, pointing to the apparent contradiction between Coxon’s warnings that AI could kill everyone and his decision to keep profiting from one of the companies he says is racing towards that outcome. “Do you prefer dying to liking money?” Reich asked.

Coxon argued that giving away the equity would be a symbolic gesture rather than a meaningful one. “I don’t think it’s gonna… If I thought for a moment that donating it, like, would change anything, then sure, I’d do it,” he said.

Reich wasn’t satisfied, suggesting that donating the stock “might give you more moral credibility, though.”

“I agree,” Coxon said, “but you could also stop asking the questions.”

From “Gambling With Our Lives” To “I Like Money, Bro”

The comments are a sharp turn from the tone of Coxon’s resignation in September. The 27-year-old, who spent three years on pretraining research first at OpenAI and then at Anthropic, quit with a viral thread claiming that the people building frontier AI privately believe it could kill everyone by the end of the decade, and that neither of his former employers was acting responsibly given that belief.

The post racked up well over a hundred million views and set off one of the most intense weeks of AI safety discourse in recent memory. Anthropic’s Alignment Science lead Evan Hubinger publicly backed Coxon and put the odds of AI killing all humans above 10 percent within a decade. Anthropic CEO Dario Amodei said he agreed with Coxon “much more” than he disagreed with him, and days later published his essay arguing that AI development must slow down to “pace the frontier” — a call that Sam Altman and Elon Musk quickly endorsed.

The backlash was just as loud. NVIDIA CEO Jensen Huang called Coxon’s comments outlandish and “deeply untrue”, David Sacks described his virality as a “psyop”, and President Trump went on to accuse Amodei of pretending to be a “perfect little angel” by asking for regulation.

Anthropic Equity Gone, OpenAI Equity Kept

Part of what made Coxon’s resignation so compelling at the time was the apparent personal cost. He told Axios that he had left Anthropic roughly two months before his equity there would have vested, meaning he walked away from that stake entirely.

But his OpenAI equity was a different matter, and it’s likely worth a great deal. OpenAI is now the most valuable AI startup in the world at a reported $1.25 trillion, and stakes in frontier labs have become so valuable that sellers are listing San Francisco homes in exchange for OpenAI or Anthropic stock. Coxon didn’t say how much his stake is worth.

The admission also comes as Coxon appears to be rethinking his departure more broadly. At the same Stanford event, he reportedly said he wasn’t sure he’d made the right decision in leaving, and floated the possibility that staying on to do safety research might have been more effective — even suggesting he might go back.

Honesty Or Hypocrisy?

There are two ways to read Coxon’s comments. The charitable reading is that he is simply being unusually honest. He has argued that individual researchers resigning won’t stop AI development without international coordination, and by the same logic, donating his OpenAI shares wouldn’t slow down OpenAI by a single day. He has also never claimed to be a saint, and few people in his position would give up what could be a life-changing sum of money for a gesture.

The less charitable reading is that it undercuts the very thing that made his warning land. Coxon’s resignation went viral because it read as a costly signal from someone on the inside willing to sacrifice something for his beliefs. Holding on to equity in a company he says is gambling with humanity’s survival — while openly calling it blood money — hands his critics exactly the ammunition they’ve been looking for. Those who argue that AI doom warnings are driven by self-interest now have a clip of the industry’s most famous doomer saying “I like money, bro.”

And Coxon’s parting shot to his interviewer — that his moral credibility would be just fine if people simply stopped asking — is unlikely to make the questions go away.

Posted in AI