Sam Altman is no longer just nodding along from the sidelines. The OpenAI chief executive says he agrees with Anthropic CEO Dario Amodei's call for the AI industry to deliberately slow its pace of capability gains, marking one of the most direct public alignments between the two rival lab heads in the generative-AI race so far.
Amodei laid out the case in a roughly 3,800-word essay published over the weekend, arguing the industry needs to “pace the frontier” rather than race unchecked toward more powerful models. “We must slow the pace at which we improve the capabilities of AI models,” he wrote. “Progress will still seem fast, and we must make wise use of the time we gain.” Altman responded in a post of his own, writing that he agrees with Amodei “that we need to pace the frontier,” adding that the idea had already been “a primary topic of discussions” inside OpenAI in recent weeks. He also endorsed a specific proposal from Amodei's essay: giving independent evaluators employee-like access to frontier models before they ship, saying OpenAI “will do the same.”
The rare display of unity between two companies that compete fiercely for talent, compute and enterprise customers comes as regulators and safety researchers have grown louder about the risks of an unchecked capability race. Fortune reported that Altman, in the same interview, hinted at the possibility of a broader safety pact among leading AI labs, though he declined to detail any private discussions already underway, telling the outlet only that formal coordination between the major labs is something he expects to happen.
Altman used the same conversation to rule out an OpenAI initial public offering in the near term. Asked whether the company might go public in 2026 or 2027, he said “not 2026,” adding that the safety work still underway makes this “an ill-advised moment to go public” and that OpenAI is “not rushing into” a listing. The comments land at a moment when rival Anthropic is reportedly in early discussions of its own that could value the company near $2 trillion in a future listing, underscoring how differently the two labs are approaching the public markets even as their CEOs converge on safety strategy.
Related: Jensen Huang Declares 'AGI Has Arrived' as GPT-6 Astra Benchmarks Land
Whether “pacing the frontier” changes anything in practice remains to be seen. Neither Amodei's essay nor Altman's response commits either company to a concrete timeline, benchmark, or enforcement mechanism — the substance is a shared acknowledgment that capability gains have been outrunning safety evaluation, not a plan to formally throttle model releases. Regulatory scrutiny of the industry's biggest deals is already intensifying regardless: the Justice Department is investigating whether Nvidia's $20 billion deal for Groq dodged antitrust review, a sign that the safety conversation among lab leaders is unfolding alongside separate pressure from antitrust enforcers.
The moment also comes as capital keeps pouring into the sector at scale. Nvidia is reportedly in talks to anchor Anthropic's own record-setting IPO with a $10 billion commitment, a deal that would make Altman's rival the far more aggressive of the two labs on public-market timing, even as its CEO leads the call to slow down the underlying technology race.
FAQ
What does “pacing the frontier” actually mean?
It refers to Dario Amodei's proposal that AI labs deliberately slow the rate at which they increase model capabilities, giving safety evaluation and independent oversight time to catch up, rather than a specific technical limit or regulatory rule.
Did Sam Altman commit OpenAI to any concrete changes?
He committed to one specific step — giving independent evaluators employee-like access to frontier models before release — but did not lay out a timeline or enforcement mechanism for slowing capability gains more broadly.
Is OpenAI going public in 2026?
No. Altman told Fortune “not 2026,” citing ongoing safety work, and said the company is not rushing toward an IPO in 2027 either.
How does this compare to Anthropic's own IPO plans?
Anthropic is reportedly in early talks for a listing that could value the company near $2 trillion, with Nvidia said to be considering a $10 billion anchor investment — a far more advanced timeline than OpenAI's.
