Amodei and Altman Converge on Slowing AI While Both Labs Keep Shipping
Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman agree that AI development needs to slow down.
That agreement was reported on September 15, 2026 Associated Press. It puts the heads of two frontier labs on the same side of a capabilities question, even as both labs continue to ship models, APIs and enterprise products.
Amodei made the more explicit public case. He called on AI companies to "slow the rate at which they advance model capabilities" Reuters. He then wrote an essay calling for slowing down AI development, according to September 14 reporting. He also urged regulations requiring "independent audits" for the AI industry.
The political response was swift and split. U.S. President Trump dismissed AI safety alarms in September 2026, saying the U.S. "already has tools to police the AI industry" Reuters. A China state newspaper criticized Anthropic's calls to slow AI development as a "Cold War tactic" Reuters.
OpenAI has its own safety language on the record. The company publishes an official page titled 'An Alien Mind' on its own domain. That page states the time calls for "extreme caution" and that "no one is prepared for the consequences of a continued rapid rise in machines."
OpenAI also publishes an official page titled 'AI progress and recommendations.' That page, dated November 6, 2025, states OpenAI is "deeply committed to safety."
Those statements now sit alongside a dense product schedule. On September 16, 2026, OpenAI published 'Reimagining advertising with AI,' 'How to connect AI usage to business value' and 'Our framework for reporting model misalignment.'
The prior week was equally busy. On September 11, OpenAI published 'Rapidly scaling online storage to serve over 1 billion ChatGPT users.' On September 10, it published 'How a researcher uses Codex and ChatGPT to search for new antimicrobial molecules,' 'Now everyone can put data to work,' 'Introducing ChatGPT for Financial Services,' 'Build more natural voice experiences with GPT-Live-1 in the API' and 'Introducing the Agents API.'
Anthropic is also shipping. On September 1, 2026, it announced Claude Fable 5.1 and Claude Mythos 5.1. The company describes the pair as its "most advanced models for coding and knowledge work."
For practitioners, the distinction matters. Slowing capability advances is not the same as slowing product releases. Capability work concerns pre-training scale, new reasoning behavior, autonomy and evaluation gaps. Product work concerns packaging, retrieval, voice latency, agents orchestration, storage and vertical compliance. The September publications from OpenAI fall largely in the second bucket. Agents API, GPT-Live-1 for voice, financial-services controls and methods for tying usage to business value are deployment and integration problems.
The broader context here is a familiar industry split between model risk and deployment risk. Amodei's audit proposal addresses the first. It implies third-party access, repeatable evaluations and disclosure obligations that would apply across labs. Trump's response addresses the second view, that existing oversight is sufficient. Beijing's response treats the slowdown call as geopolitical rather than technical.
In this author's view, worth flagging is how closely the slowdown language tracks with scale. A lab serving over 1 billion ChatGPT users and operating research agents for antimicrobial discovery is no longer talking about hypothetical misuse. Misalignment reporting, independent audits and storage infrastructure are operational concerns at that size. My kids grew up with search, then smartphones, then generative chat, and each transition looked gradual until support systems had to be rebuilt quickly. Enterprise AI feels similar now.
That is also why the current moment is constructive. Reporting frameworks for model misalignment, audit requirements, clearer business-value measurement and domain-specific controls for financial services are the unglamorous pieces that let powerful systems be used routinely. If labs can agree that capabilities need restraint while competing on safety tooling, reliability and verifiable deployment practices, technology professionals gain something usable. Caution, in that sense, enables adoption rather than blocking it.
