Newsletters
OpenAI said Tuesday it is working with Anthropic and Google to develop an industry-wide standards body for artificial intelligence, an effort aimed at addressing mounting concerns over the technology's risks even as rival companies clash over who should set the rules.
The initiative traces back to a proposal from Demis Hassabis, head of Google DeepMind, who called in July for a self-regulatory organization modeled on the Financial Industry Regulatory Authority, the body that oversees brokers and investment firms in the United States. Hassabis argued the group should test the most powerful AI systems before they reach the public.
In his July statement, Hassabis said the effort would require significant financial backing largely drawn from AI companies themselves. "Funding would need to be substantial and likely mostly come from industry, in order to attract world-class technical talent and provide the necessary compute resources for large-scale testing," he wrote.
While OpenAI, Anthropic and Google appear to be aligning behind the concept, there is no broader consensus across the industry on how such oversight should be structured or implemented.
The plan has already drawn sharp criticism from outside the three companies involved. Aidan Gomez, co-founder and chief executive of Canadian AI company Cohere, accused the dominant American AI labs of using safety as cover to consolidate control over the industry.
In a blog post published Sunday, Gomez wrote that guardrails for AI are not in dispute, but that the fight is over authority. "The dispute is over who writes them, who gets to participate and whose interests the rules are protecting," he wrote, describing the arrangement as a "cartel."
The push for coordinated oversight comes amid a string of public warnings from AI researchers who have recently left major labs. Jacob Coxon, who departed Anthropic last week after previously working at OpenAI, said AI companies were "gambling with our lives."
On Monday, former Google DeepMind researcher Bilal Chughtai voiced similar alarm, writing on social media, "I earnestly believe that AI has the potential to kill us all, and that we might be running out of time to avoid this outcome."
Anthropic chief executive Dario Amodei called Saturday for a slowdown in the pace of AI development, a position that drew public support from OpenAI's Sam Altman, Elon Musk and Hassabis.
Not everyone shares that sense of urgency. President Donald Trump has dismissed warnings about AI risk as "hoaxes." Nvidia chief executive Jensen Huang also rejected the doomsday framing on Monday at a Los Angeles conference, where Trump phoned in while Huang was speaking.
Huang said predictions that AI could destroy humanity are "not grounded on science," adding, "The confluence of these words and then the prediction is alarming and troubling and it shouldn't be done. It's irresponsible."
The disagreement underscores a widening divide within the AI industry, one pitting calls for coordinated, industry-led oversight against accusations that such efforts serve to entrench the power of a handful of dominant firms, even as political leaders and executives remain split on how seriously to treat warnings of catastrophic risk.
Set as preferred source
Set as preferred source
1 min read
1 min read
