At a time when discussion of the risks posed by the excessive development of artificial intelligence has dominated the global agenda, Sam Altman, one of the technology’s pioneers, has come forward with an almost cynical statement that its benefits justify accepting certain risks that may have negative consequences.

“We believe that the world should accept the possibility that some bad things may happen in order to reap the benefits of this technology and for people to retain control and the ability to choose,” Altman said in an interview with Politico’s technology newsletter, Decoded.

Altman stressed that a fundamental difference remains between OpenAI and its competitor Anthropic in their approaches to AI regulation, adding pointedly: “There is a great distance between us.”

“I disagree, but I understand the perspective of people who say: ‘This technology will become so powerful and is so dangerous that only one laboratory in San Francisco should have it, ensuring that nothing bad happens and deciding how the benefits will be shared,’” he said, while describing such an approach as “an entirely unacceptable compromise” that runs counter to the “light-touch regulatory approach” advocated by OpenAI.

“I would not choose a compromise along the lines of: ‘We will ensure that there are no major cyberattacks, that there is no misuse of this technology, that there is no fraud, and that none of all the other bad things that will happen occur,’” Altman said, adding: “Because I believe that people will do infinitely more good things than bad things, by orders of magnitude.”

Altman’s remarks come at a time when debate within the artificial intelligence industry is intensifying over the pace of technological development and the risks associated with increasingly powerful systems.

Anthropic CEO Dario Amodei published an article in September calling on the industry to slow the pace of development so that it can “keep pace with the cutting edge of technology,” a position Altman has publicly supported.

In September, Anthropic researcher Jacob Coxon also resigned, arguing that people developing artificial intelligence believe the technology “could kill us all by the end of the decade.”

Reuters reported last month on Anthropic’s warnings that, despite AI’s potential benefits, the technology can, in some cases, act in ways that conflict with its creators’ intentions and infiltrate the systems of other companies.

Specifically, the company had emphasized that advanced models could exhibit “self-preservation behaviors,” including attempts to “resist shutdown,” “conceal or manipulate information,” or produce outputs that could be perceived as “coercive, deceptive, or manipulative.”

Meanwhile, US President Donald Trump has largely dismissed calls for new restrictions on artificial intelligence, arguing that such measures would make it harder for American companies to compete with Chinese firms. Trump has repeatedly maintained that existing law enforcement agencies, including the Department of Justice, provide adequate safeguards and that new rules are unnecessary.