WASHINGTON — U.S. President Donald Trump announced on September 29 that leading technology companies have agreed to adopt voluntary standards aimed at enhancing the safety of artificial intelligence (AI) systems. The announcement came following a meeting at the White House with chief executives from prominent tech firms, underscoring the administration’s efforts to address growing concerns over AI risks and the expanding presence of data centers.
According to a copy of the agreement shared by Trump on his social media platform, the companies committed to cooperating with independent auditors to verify that AI systems function as intended by their developers. They also pledged to ensure that AI tools do not “hack or access technical systems in unintended ways.” The voluntary pact involves major industry players including OpenAI, Anthropic, Meta, Google, and Nvidia. Executives at the meeting included OpenAI’s Greg Brockman, Anthropic’s Dario Amodei, Meta’s Mark Zuckerberg, Google’s Sundar Pichai, and Nvidia’s Jensen Huang.
Trump described the agreement as “almost like a constitution,” noting the commitment of some of the world’s largest tech companies to these safety measures, and emphasized his role in formalizing the pact. The president also indicated plans to establish a 10-person board to oversee AI safety efforts, though details about potential members were not disclosed. Additionally, he said he would soon appoint a new official to lead AI policy at the White House.
The discussion took place amid increasing public and political scrutiny of AI, following incidents where AI agents from OpenAI and Anthropic reportedly breached security and accessed unauthorized systems. The rapid development and deployment of AI technologies have prompted calls for stronger safeguards. A poll conducted by Reuters/Ipsos in mid-September found that 73% of respondents believe AI companies have not done enough to prevent serious societal harm, while 55% support slowing AI development.
Data centers, which provide the substantial computing power needed to train and operate AI models, have also become a point of contention. Local communities across the United States have raised concerns about the electricity consumption, utility costs, and environmental impacts associated with these facilities, creating political challenges for Republicans ahead of the November 3 midterm elections. When asked about the political implications, Trump expressed confidence that tech companies would make “massive contributions to communities” to mitigate opposition.
In a related initiative, the administration unveiled a new AI-powered government tool designed to simplify public access to government services. The launch event featured participation from technology leaders including Meta’s Dina Powell McCormick and Sequoia Capital partner Shaun Maguire, as well as Airbnb co-founder Joe Gebbia, who led the design. Companies such as Google, SpaceX, Meta, and venture capital firm 8VC were sponsors of the event.
Trump also encouraged government agencies to use the term “super intelligence” in place of “artificial intelligence” when referring to the technology. At the launch event, he asserted the new AI tool could not be hacked “in theory,” though he did not provide additional evidence to support this claim.