Mark Zuckerberg, the chief executive of Meta Platforms, stated on Tuesday that competitive pressures and legal liabilities are sufficient to compel artificial intelligence companies to adopt safety measures on their own. This position appears to diverge from recent appeals by leaders of several top AI firms who have called for a coordinated slowdown in AI development.

His comments highlight a growing divide among industry executives and US authorities over how to address warnings about AI risks. Some insiders caution that the technology is advancing faster than companies can manage, potentially posing an existential threat to humanity.

In a post on the social platform X on Tuesday, Zuckerberg asserted that every AI lab has both the responsibility and the incentive to progress at a pace that allows for safe model training, and has the capability to act independently to achieve this goal.

Just three days prior, Dario Amodei, the chief executive of Anthropic, urged AI companies to slow the acceleration of their models' capabilities. This followed warnings from several researchers that AI could threaten human survival. Sam Altman, CEO of competitor OpenAI, and Elon Musk of xAI both publicly endorsed Amodei's call within hours.

Zuckerberg argued on Tuesday that trust and alignment are rapidly becoming the most critical differentiators among AI agents and models. He stated that any lab failing to prioritize alignment will fall behind its competitors. He also noted that labs have strong incentives to prevent their models from causing harm, as they would otherwise face "significant" legal consequences.

As an example, he cited Meta's decision earlier this year to postpone the release of its Muse AI agent to strengthen its safety features. "We didn't ask other companies to go first before we acted," Zuckerberg wrote. "We simply treated this as part of our routine work, because it is clearly in the best interest of both our users and our company."