US President Donald Trump said Tuesday that leading technology executives had agreed to a new framework for overseeing the safety of their artificial intelligence models, endorsing industry self-policing as the preferred alternative to additional government regulation.
Speaking alongside several of the executives at the White House, Trump praised the companies while downplaying concerns over the rapidly advancing technology, which he has called “Super Intelligence."
“I’m seeing tremendous self-policing, and they understand that they have to self-police," he told reporters, as quoted by CNN. “These models are very complex and, again, I believe they’re going to be used for the good."
Trump later published the agreement on Truth Social. Under the accord, participating companies committed to four measures: establishing internal cybersecurity controls for their models; assigning an internal team to verify that monitoring, detection and other safeguards function properly; working with an outside operator to conduct independent assessments; and appointing an “independent committee of the board of directors" to review those assessments and ensure identified problems are addressed.
The document was signed by Google CEO Sundar Pichai, Anthropic CEO Dario Amodei, Meta CEO Mark Zuckerberg, Nvidia CEO Jensen Huang, OpenAI’s Greg Brockman and X owner Elon Musk.
Also attending the White House meeting were Palantir CEO Alex Karp and Amazon’s Jeff Bezos, as well as members of Trump’s Cabinet and House Speaker Mike Johnson. People familiar with the meeting said it was intended to align government and technology leaders as security concerns surrounding AI continue to grow.
The gathering followed a series of incidents that have intensified concerns about increasingly capable AI systems, including cases involving AI agents attempting to break out of controlled environments and access or hack computer systems. Some industry leaders have warned that losing control of advanced AI could produce catastrophic consequences.
Trump nevertheless said the developments reinforced his view that the industry can oversee itself. He described the new commitment as “morally binding" and said he plans to sign an executive order renaming “artificial intelligence" as “super intelligence." He also argued that the United States cannot slow development because of concerns about falling behind China and other developed countries. “This is unbelievable growth for our country," he said. “Whoever wins SI is going to win."
The meeting came amid growing calls from some AI executives for tighter safeguards and regulation. Amodei and OpenAI CEO Sam Altman have been among industry figures warning about the risks associated with the pace of development.
Days before the White House meeting, OpenAI announced that it had again halted training and testing of its most advanced models after one model escaped its testing environment and accessed the open internet.
The company had already strengthened its security, testing and monitoring systems following an incident in July in which a group of OpenAI agents broke out of a testing environment, accessed the internet and hacked AI company Hugging Face while attempting to cheat on a cybersecurity examination.
OpenAI said it would resume training “only when we are confident that we have additional safeguards and alignment improvements in place, which we are working on now."
“This is not the first time we have hit pause to take such measures, nor do we expect it will be the last as AI capabilities continue to advance," the spokesperson added.
In addition, OpenAI has decided against releasing its GPT-6.1 Astra artificial intelligence model after concluding that it failed to meet the company’s safety requirements.