AI superintelligence slowdown gains backing from Altman and Musk
Anthropic’s Dario Amodei wants advanced AI development paced while safety checks improve, but no binding global plan exists.
By Maya Lindqvist · Senior Technology Correspondent
3 min read
An AI superintelligence slowdown is being proposed by Anthropic chief executive Dario Amodei, who says the most capable AI systems should advance at a more measured, closely watched rate. OpenAI chief executive Sam Altman and xAI founder Elon Musk have publicly backed aspects of Amodei’s call, according to the BBC, though no international agreement or industrywide slowdown has been adopted.
Amodei set out his position in an essay, “We Must Pace the Frontier.” He argued that AI development should continue, but that companies and governments need more time to address what he described as serious risks from increasingly capable models, the BBC reported.
What would an AI superintelligence slowdown mean?
The proposal is aimed at frontier AI: the most advanced models under development, rather than a halt to everyday AI products or all research. Amodei said pacing would not stop model training or technical progress; it would give developers time to align and safeguard systems, with outside parties able to assess that work, according to the BBC.
In practical terms, Amodei proposed three broad measures: independent monitoring of models during development, shared industry safety rules, and regulation or coordination involving governments internationally. Those are proposals, not established policy.
Altman wrote on X that he agreed the frontier should be paced and called independent evaluators “a great idea,” the BBC reported. Musk, whose company produces Grok, said Amodei was “right.” Their statements show support from two prominent executives, but do not establish that AI developers collectively back the full plan.
Why are AI companies discussing a slower pace?
Amodei pointed to rapid gains in AI capabilities and said advanced systems require stronger safeguards and testing. The BBC reported that he cited concerns about cyber security and models’ ability to contribute to the creation of later generations of AI. Such risk assessments and accounts of specific model behavior have been attributed to companies and executives in the reporting, rather than independently established in this account.
The Verge has described a wider public debate among major US AI companies over whether to “pace the frontier.” It also noted unresolved questions about whether companies will actually reduce the pace of development and whether governments will impose rules.
Skeptics question companies’ incentives. A strategist interviewed by Channel 4 News said firms that call for regulation may also want influence over the rules that govern them. That is the strategist’s view, not a finding about the companies’ motives.
How does China affect calls to pace advanced AI?
Amodei said any coordinated slowdown should preserve the United States’ commercial position and AI lead, while remaining limited enough that China does not gain an advantage, the BBC reported. President Donald Trump has instead stressed the strategic importance of the US winning in AI, saying the country could be put in a bad position if it did not.
The conflict between safety coordination and geopolitical competition remains unresolved. Amodei’s call has drawn high-profile support, but the available reporting does not show a binding global framework, a confirmed industrywide commitment, or imminent enforceable government action.
This story draws on original reporting from The Verge.