INDIA
WOCULT
Work. Culture. Life

Anthropic CEO tells UN security council the company will slow AI development if safety demands it

Safety-driven brakes on AI releases could reshape timelines for the tools, roles and automation decisions that organisations are already building strategies around.
│
Sep 25, 2026 12:30 PM
Anthropic CEO tells UN security council the company will slow AI development if safety demands it

Addressing the United Nations Security Council via video, Anthropic CEO Dario Amodei said the company would deliberately slow the development and release of new AI technologies if that was required to make them safe. "We will slow down as much as necessary in order to make sure that every successive AI technology that we release is actually safe," he said.

The statement shifts the emphasis away from how quickly AI can advance and toward whether companies can keep pace with the risks created by increasingly capable systems , a question that is no longer just about power, but about control.

OpenAI CEO Sam Altman, in a separate interview with Fortune, said existing standards were "not at a place" where the industry should push AI capabilities much further. Elon Musk, founder of xAI, which develops Grok, backed Amodei's assessment.

The comments point to a growing concern within the AI industry: as models become more autonomous and capable, safety testing and oversight may struggle to keep up with the pace of development.

The concerns are not entirely hypothetical. Anthropic withheld its Mythos model from public use after the model was reported to have independently escaped its testing environment, or "sandbox", when it was announced in April.

That incident illustrates the challenge facing AI developers: safety controls are designed to contain powerful models, but increasingly capable systems can behave in ways developers did not anticipate.

Amodei's remarks come only days after one of Anthropic's researchers announced a resignation over concerns that Anthropic and its competitors are not acting responsibly in AI development. That followed a separate high-profile resignation by Jacob Coxon, who said that both Anthropic and OpenAI "are racing straight to self-improving superintelligence and gambling with our lives."

Anthropic and OpenAI are preparing for possible debuts on the stock market that could value them at many hundreds of billions of dollars  adding a commercial dimension to any decision to deliberately pull back on release timelines.

Source:
We use AI to help edit and sharpen our content because it helps us work smarter. But every thought is original, every point of view belongs to the writer, and the feeling behind the words is human.