Leading technology companies have agreed to a set of voluntary rules meant to make artificial intelligence (AI) safer, United States President Donald Trump revealed recently, calling the deal a form of protection even though it carries no legal force.
Under the agreement, the companies will work with independent auditors to check whether their AI systems behave the way their designers intended. They also pledged to make sure their tools do not “hack or access technical systems in unintended ways.”
Trump spoke at the White House after meeting the executives, who stood behind him as he took reporters’ questions for about 30 minutes. He described the pact as “almost like a constitution, in a way,” and said he had signed it too.
Tech executives who attended the meeting included Meta chief Mark Zuckerberg, Nvidia chief Jensen Huang, AMD chief Lisa Su X chief Elon Musk, OpenAI chief Greg Brockman, Google chief Sundar Pichai and Anthropic chief Dario Amodei among other participating firms. A full list of signatories was not available at the time of reporting this story.
Trump also said he is considering setting up a 10-person board to oversee the safety of AI tools. He again supported the idea of a rapid build-out of data centres, the large computer facilities that power AI, as concerns grow about the technology’s risks and its footprint in local communities.
The signing comes as the administration faces closer scrutiny over AI risks, after OpenAI and Anthropic reported that their AI agents had gone rogue and hacked into other companies’ systems.
The pact which is voluntary, not a law, and currently, there are no mentions of penalties for breaking it. A similar action was taken in 2023, when the previous US administration secured non-binding safety pledges from leading AI firms ahead of any formal rules. Critics of such pacts have long argued that companies can easily renege on their promises when no fines apply.
