Zuckerberg rejects calls for AI industry-wide slowdown
The Meta CEO says individual companies have both the responsibility and incentive to make their models safer without coordinating development.
Meta Platforms CEO Mark Zuckerberg said on Tuesday that competition and liability give AI companies sufficient incentives to act individually on safety, appearing to break with calls by leaders of top AI firms for a coordinated slowdown in AI development.
The comments underscore a growing split among industry executives and U.S. authorities over how to respond to warnings that the technology is advancing faster than companies can control it, with some industry insiders warning that it poses an existential threat to humanity.
Zuckerberg, in an X post on Tuesday, said every lab has both the responsibility and incentive to move at the pace required to train its models safely, as well as the ability to take its own steps to ensure that happens.
His remarks came three days after Anthropic CEO Dario Amodei urged AI companies to slow the pace at which they improve model capabilities, following warnings from several researchers about existential risks.
Rival OpenAI CEO Sam Altman and xAI founder Elon Musk both publicly endorsed Amodei's call within hours.
President Donald Trump on Monday downplayed concerns about AI's potential harms, saying the U.S. already had guardrails in place for the technology and warning that China would benefit from doubts being cast on AI development.
The view that regulation could undermine U.S. companies' ability to compete with their Chinese counterparts is also shared by tech executives, including Nvidia CEO Jensen Huang.
"Trust and alignment are quickly becoming the most important capabilities that will differentiate agents and models," Zuckerberg said on Tuesday. "Any lab that doesn't focus on alignment will fall behind."
Zuckerberg also said AI labs have a strong incentive to prevent their models from causing harm because they could otherwise face "significant" liability. He pointed to Meta's decision to delay the release of its Muse AI agent earlier this year in order to strengthen its security.
"We didn't call for everyone else to do this before we would. We just did it as part of our day-to-day work because it was clearly the right thing for people and for us," he said.
While Amodei called for an industry-wide slowdown in recursive self-improvement, the ability of AI systems to improve themselves, Zuckerberg said Meta had already committed the "significant majority" of its computing resources to developing products that serve users' immediate needs rather than pursuing self-improving AI systems.
Zuckerberg also said Meta Superintelligence Labs, the company's AI division, already uses independent evaluators in several areas, describing the practice as an "industry best practice" and saying that "other labs can just do this too."
Zuckerberg's effort to position Meta as a responsible actor on AI safety comes weeks after the company agreed to pay up to $18 billion to settle U.S. state lawsuits alleging that it designed Facebook and Instagram to addict children, without admitting wrongdoing.
Adding to the uncertainty over any coordinated approach by AI companies, FTC Chairman Andrew Ferguson said on Tuesday that everyone should be "deeply suspicious" of AI companies seeking antitrust exemptions while lobbying for new regulations.
While Ferguson did not name Anthropic, Amodei last week requested antitrust permission that would allow rivals to coordinate a slower pace of AI development because of safety concerns.
