OpenAI CEO Sam Altman has said his company and rival Anthropic remain divided by a basic difference in outlook on how AI should be regulated. He argued that the technology's upside justifies tolerating some risk and that the public should keep broad access to it.
Altman made the remarks in an interview with Politico's Decoded newsletter. Asked where he parts ways with Anthropic and its CEO Dario Amodei, Altman said there is 'a lot of daylight' between them.
"The world should accept some bad things"
Altman said OpenAI's view is that society has to live with some negative outcomes in exchange for the technology's benefits and for people having agency over how they use it.
He rejected the idea of trading away that openness for guarantees of safety. By his account, he would not accept a bargain that promised no major hacks, no misuse and no scams if the cost was restricting the technology, because he expects the good done with AI to outweigh the harm by orders of magnitude.
At the same time, Altman said he does not accept the most extreme risks, including a serious loss of control over AI systems, according to the report.
Pushing back on the 'single lab' argument
Altman also addressed the regulatory-capture charge that tech advisers close to President Donald Trump have aimed at Anthropic. That criticism holds that safety-focused rules could entrench big incumbents.
Altman said he disagrees with the view that AI is so powerful and dangerous that one San Francisco lab should control it and decide how its benefits are distributed, though he said he understands why people hold it. He called such a trade-off "completely unacceptable" and contrary to OpenAI's lighter-touch approach to regulation.
Positions have been converging
Anthropic has long been seen as the AI industry's most vocal supporter of regulation, while OpenAI favored a lighter touch. According to the report, OpenAI has recently backed tougher rules as it deals with high-profile hacking incidents and rogue behavior by its models.
