President Donald Trump hosted a meeting at the White House on September 29, 2026, with leaders from Meta, OpenAI, Nvidia, Google, and Microsoft. The participants signed a voluntary agreement focused on AI safety, which the administration described as a morally binding framework.
The accord emphasizes self-regulation, requiring companies to conduct internal and external audits of frontier AI models. These protocols aim to identify dangerous capabilities both before and after the deployment of advanced systems, allowing firms to manage oversight without government-imposed mandates.
During the event, Trump proposed the formation of a ten-person oversight committee to guide future AI regulations. He also suggested that the industry replace the term artificial intelligence with the phrase super intelligence.
This agreement builds upon previous voluntary commitments made by tech companies since 2023, including pledges from July 2023 regarding watermarking content and sharing safety research. The current approach contrasts with the regulatory model adopted by the European Union.
