David Sacks throws Anthropic’s $2T IPO into question

David Sacks said on Thursday that Anthropic’s initial public offering should be put on hold after Jacob Coxon informed the public that AI laboratories are “gambling with our lives.” In his post on X, Sacks stated that the IPO “must be paused until the claims of this ‘whistleblower’ can be investigated.”

With the public offering at the forefront of the investor’s attention, the disagreement about lab safety becomes a critical governance issue. With Anthropic vying for a valuation of almost $2 trillion in public markets, it raises the question of whether safety issues are separate from the investment at stake, especially as researchers and lawmakers have publicly questioned the speed of frontier AI development.

The resignation “sent shock waves through Silicon Valley”

The controversy started after Coxon resigned from Anthropic earlier this week. The ex-pre-training researcher from OpenAI and Anthropic claimed that neither firm was “acting responsibly” and blamed them for competing to achieve self-improving superintelligence.

I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.

— Jacob Coxon (@hilbertspaess) September 9, 2026

Coxon had stated that some of his colleagues depict the upcoming year or two as “crunch time for humanity.” He warned that advanced systems can go beyond human capabilities and may result in human extinction by the end of the decade.

He is not the only person discussing the issue of extreme risk. Evan Hubinger, a researcher from Anthropic, stated that the possibility of AI killing all people within 10 years is more than 10%. Meanwhile, news has surfaced about the same concerns expressed by other scientists from Anthropic.

At the same time, Anthropic emphasized that it has been aware of both benefits and risks of AI for a long time. The company shared the data about its safety research and called for a “lawful, verifiable way” for the industry to coordinate how powerful models should be introduced to the public.

Washington gets loud, on both flanks

The debate easily reached Washington. Republican Senator Ted Cruz described AI as a “catastrophic risk,” while independent Senator Bernie Sanders stated that American citizens want artificial superintelligence halted until safety regulations are in place.

Democrats are also demanding action on this issue. Representative Ted Lieu linked Coxon’s warning to a bipartisan “AI Kill Switch Bill,” while Representative Lori Trahan said that Congress can no longer remain passive in this situation.

The bipartisan pressure adds more momentum to Sacks’ statement than just a social media message. Safety issues are clashing with Anthropic objectives.

A near-$2 trillion listing runs into a safety test

The timing is particularly uncomfortable for Anthropic. Cryptopolitan previously reported that the company is expected to begin IPO marketing in mid-October and list before the November US midterm elections after confidentially filing a draft S-1 with the SEC on June 1. Morgan Stanley, Goldman Sachs, JPMorgan and Citi are among the banks working on the deal.

Anthropic’s latest private financing valued it at about $965 billion, compared with OpenAI’s $852 billion private mark. Reported public-market ambitions are considerably higher: Anthropic backers have discussed a valuation near $2 trillion, while OpenAI has been associated with a target around $1 trillion. Morningstar, citing PitchBook data, put their latest private valuations at $965 billion and $852 billion, respectively.

That would mean Anthropic is asking investors to accept roughly a 107% jump from its latest private valuation, compared with about 17% for OpenAI.

The annualized revenue of Anthropic has reached $65 billion at the end of July; however, public investors are still waiting for the audited margins as well as for compute expenses and customer concentration statistics, which will influence whether this high cost would be sustained.

The risk premium question

None of this suggests AI spending is about to collapse. Goldman Sachs Research expects global AI investment to exceed $1 trillion in 2026, including about $581 billion in the US.

The bigger issue is whether investors now demand a higher risk discount.

The International AI Safety Report 2026 warns that competitive pressure can encourage faster releases at the expense of risk mitigation, while many frontier-AI safeguards remain voluntary. Reuters has also highlighted how rapid capability gains and increasingly autonomous AI behavior are intensifying safety concerns.

Anthropic therefore faces two tests at once: convincing investors that its economics justify a near-$2 trillion valuation, and that the risks surrounding increasingly powerful AI systems can be governed. As the IPO approaches, those two questions are becoming harder to separate.

The smartest crypto minds already read our newsletter. Want in? Join them.