Meta CEO Mark Zuckerberg has stated that artificial intelligence laboratories possess strong incentives to prevent their models from causing harm, appearing to diverge from calls for a global slowdown in AI development led by figures like Anthropic CEO Dario Amodei.
In a post on X, formerly Twitter, Zuckerberg asserted that "every lab has the responsibility and incentive to move at the pace required to train its models safely, and the ability to take its own actions to ensure that happens." He further suggested that trust and alignment are becoming critical differentiators for AI models, and companies that neglect these aspects risk falling behind.
Zuckerberg's remarks come amid a growing debate about the potential risks associated with rapidly advancing AI technology. Amodei, in a recent essay, argued for a deliberate pacing of AI capabilities advancement to allow risk prevention measures to keep pace. This viewpoint has been echoed by leaders such as OpenAI CEO Sam Altman, xAI CEO Elon Musk, and Google DeepMind Chair Demis Hassabis.
However, Zuckerberg countered this perspective, emphasizing that companies are already motivated to prioritize safety due to potential liability. He noted that Meta itself postponed the release of its AI agent, Muse, to ensure its safety and security, but did not advocate for a broad pause across the industry.
Nvidia CEO Jensen Huang also weighed in on the matter, suggesting that innovation and safety are not mutually exclusive. At a conference hosted by Salesforce, Huang advised AI developers to proceed rapidly but to pause and ensure safety if a product becomes uncontrollable or unsafe.
Former President Donald Trump has previously dismissed concerns about AI, calling them a "hoax" and stating that the idea of robots taking over the world is not plausible. Zuckerberg has also previously expressed skepticism regarding the establishment of a national AI regulatory body, according to reports.