AI Firms Could Face ‘Significant’ Liability If Their Systems Are Unsafe, Meta CEO Mark Zuckerberg Warns

In recent days, the rapid development of highly capable AI systems has given rise to a debate over how quickly the technology should advance, as well as the need for oversight to manage its risks. Meta CEO Mark Zuckerberg has now weighed in on the discourse, arguing that AI companies have strong incentives to ensure their systems are safe. The executive also outlined Meta’s approach to AI development and safety.

Zuckerberg’s comments come shortly after Anthropic CEO Dario Amodei called for AI companies to slow the pace of improvements to model capabilities, particularly recursive self-improvement. OpenAI CEO Sam Altman and xAI CEO Elon Musk had subsequently expressed support for Amodei’s position.

In an X post, the Meta CEO wrote about a different approach. Zuckerberg argued that AI companies do not necessarily need to wait for regulation or an industry-wide agreement to prioritise safety.

“Trust and alignment are quickly becoming the most important capabilities that will differentiate agents and models,” Zuckerberg said. He added that AI labs that do not focus on alignment could fall behind.

The executive also warned that AI developers could face “significant” liability if their systems are found to be unsafe. He pointed to Meta’s decision to delay the release of its Muse AI agent earlier this year, claiming the company had used the additional time to improve its security. The company did not wait for other AI firms to adopt similar practices and instead made the decision as part of its own development process.

Zuckerberg also said that instead of focusing on self-improving AI systems, Meta has committed the “significant majority” of its computing resources towards developing AI products aimed at addressing users’ immediate needs.

The comments arrive amid the industry facing a debate over whether companies should voluntarily slow development or whether governments should introduce additional oversight. US President Donald Trump previously played down concerns about AI risks. Meanwhile, Nvidia CEO Jensen Huang also argued that excessive regulation could affect the ability of US companies to compete with China.

Meta itself, notably, has faced scrutiny over its approach to its consumer platforms in the past. The company recently agreed to pay up to $18 billion to settle lawsuits brought by US states alleging that Facebook and Instagram were designed to be addictive to children, although Meta did not admit wrongdoing.