Musk and Tech Allies Attack Former Anthropic Researcher Over Viral AI Extinction Warning

A Viral AI Doom Warning Becomes a Political and Corporate Fight

Former Anthropic researcher Jacob Coxon set off a public storm this week when he posted a series of warnings on X accusing OpenAI and Anthropic of racing toward self-improving superintelligence and 'gambling with our lives.' One of his posts — stating that people developing AI believe the technology could 'kill us all by the end of the decade' and that this is not a marketing trick — had attracted more than 23 million views, while his posts altogether were reported to have passed 120 million views.

Elon Musk responded by questioning how such a previously nearly inactive account could generate that level of attention, and he suggested the affair could be a conspiracy or PR maneuver. Epic CEO Tim Sweeney said the posts and reactions looked 'choreographed.' Musk then said, without evidence, that the debate appeared to follow prepared groundwork for a 'psychological operation' against AI. Billionaire Bill Ackman amplified a claim that the situation looked like the start of a well-funded PR operation aimed at persuading Democrats to regulate AI 'into unusability.'

The dispute moved from social media to corporate stakes when David Sacks, a former White House adviser on AI and crypto, said Anthropic's IPO 'surely must be delayed' until the whistleblower's claims are investigated. Jason Calacanis, Sacks's All-In podcast co-host, asked whether Anthropic is running a regulatory-capture strategy or whether people inside the company are seeing something that raises their estimated probability of catastrophe. Axios later reported that Coxon worked at Anthropic for only four months and left two months before his stock options vested, while still holding equity from his earlier role at OpenAI.

Coxon has framed the two labs differently: he said OpenAI employees have not deeply internalized the civilizational stakes, while Anthropic understands the problem but is trapped in a race because it believes no one else will act responsibly. None of the most dire safety claims or the most aggressive counterclaims have been independently verified.

Why Musk, Sacks and Ackman Are Framing Coxon's Posts as a Psy-Op

Coxon's Message Is Being Tested Through Employment Details, Not Just Evidence

The core employment timeline is now part of the debate. Axios reported that Coxon spent only four months at Anthropic and left two months before his options vested, while retaining an ownership stake from OpenAI. That does not prove or disprove his warning, but it gives critics a concrete line of attack: they can argue his message is shaped by positioning, timing or affiliations rather than by new internal knowledge. In practice, the public argument has moved partly from 'is this AI risk real?' to 'what does this person have to gain?'

Musk, Sacks and Ackman Are Treating It as a Regulatory-Capture Fight

Several of the most prominent responses reframe the entire episode as a battle over regulation. Musk called the activity a possible psychological operation without supplying evidence. Ackman promoted the idea of a sophisticated PR campaign tied to Democratic support for AI regulation. Sacks has previously accused Anthropic of running a 'sophisticated regulatory-capture strategy based on fear,' and Calacanis asked directly whether Anthropic's public safety posture is designed to raise barriers to competition. That is a different dispute from whether Coxon's specific safety claims are accurate — and it is one where AI policy has become openly partisan.

Where This Leaves Anthropic and OpenAI

The immediate corporate pressure is on Anthropic. A public call to postpone its IPO, coming from a former White House AI adviser, creates a reputational issue even if no regulator has acted. Coxon's distinction between the two labs also matters: he gives Anthropic more credit for understanding the risk than he gives OpenAI, but he describes Anthropic as trapped in a race it cannot safely win. That message may reinforce customers who value a safety-first brand while also arming critics who say the safety narrative is a business strategy. The wider field is left with a public fight in which very serious safety claims are being litigated through social media reach and employment timelines rather than through published safety evidence.

What the Escalating AI Safety Clash Signals for Frontier AI Companies and Investors

For executives, investors and communications teams around frontier AI, the Coxon affair has immediate operational implications:

• Anthropic investors and board members should treat the IPO timing as an open question: David Sacks explicitly called for a delay pending an investigation, and that statement is likely to surface in due diligence and media coverage.
• AI companies with public safety positions must be ready to have those positions tested against employment tenure and equity data; Axios reported Coxon worked at Anthropic for only four months and left two months before options vested, which critics are using to attack his motives.
• Frontier model developers should prepare investor and customer questions about the same 'regulatory capture' framing that Jason Calacanis put to Anthropic, especially in sectors where AI adoption depends on trust in safety claims.
• Communications and policy teams need a documented timeline of safety reviews and public statements, because David Sacks' demand to investigate the whistleblower signals that the next conflict may be fought over audit trails rather than the underlying risk estimate itself.