Meta CEO Mark Zuckerberg has publicly set out his view on artificial intelligence safety, aligning himself more closely with the approach advocated by Nvidia chief Jensen Huang and moving away from the more restrictive stance pushed by Anthropic CEO Dario Amodei.
In a post on X and other social platforms, Zuckerberg said AI labs that fail to focus on “alignment” will fall behind. He was referring to the work developers do to ensure that the technology remains consistent with human values and can be controlled safely.
The debate intensified after Dario Amodei, in an essay published over the weekend, urged the tech industry to “slow the pace at which we improve the capabilities of AI models.” That position was later endorsed by Altman, while U.S. President Donald Trump criticized calls for such a slowdown.
“There is a lot of debate about slowing progress on capabilities until alignment catches up,” Zuckerberg said. In his view, trust and alignment are rapidly becoming the most important capabilities that will distinguish AI agents and models.
Zuckerberg argued that companies already have strong incentives to prevent problems because AI labs face “significant liability if their models cause harm.” That market-based perspective is broadly in line with Jensen Huang’s public comments that product makers should be responsible for what they release rather than relying on new restrictive regulation.
Zuckerberg also said Meta voluntarily “delayed shipping” its Muse AI technologies for “safety and security” reasons. “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.
Speaking at the annual Salesforce Dreamforce conference, Jensen Huang told Salesforce CEO Marc Benioff that AI model makers should take responsibility for their products instead of seeking new government restrictions aimed at limiting harms and risks. In a later interview on CNBC’s “Mad Money,” he described such laws and regulations as “just completely unnecessary.”
“We have plenty of laws. We have plenty of regulations that govern the reliability and the functionality of products,” Huang said.
At the same event, Dario Amodei reiterated his core argument that slowing the pace of model development is “the way to lead the industry forward, to set an example, to say that everyone can always be better.”
Later in the day, Altman joined Marc Benioff for a discussion and said safety and monitoring should take precedence over features in AI development. He also acknowledged broader fears that competitive pressure could lead some companies or countries to make the wrong choices.
