Trump, tech executives release 'morally binding' accord amid rising fears over AI safety

"It's almost like a constitution, in a way," Trump said, referring to the agreement between the companies. "And the biggest people in the world signed that, and I signed it as president."

US President Donald Trump on Tuesday said tech executives agreed to establish standards for AI and reiterated his support for the rapid expansion of data centers, as concerns mount over AI safety and the industry's growing footprint in communities.

Trump made the comments after a meeting with tech executives at the White House. The administration is facing increasing scrutiny over the risks posed by advanced artificial intelligence, after OpenAI and Anthropic reported their AI agents had gone rogue and hacked into other companies' systems.

Data centers, which provide the computing power needed to train and run AI models, have drawn opposition in communities nationwide over concerns about electricity demand, utility costs and other local impacts. The backlash poses a political challenge for Republicans heading into the US midterm elections on November 3.

"It's almost like a constitution, in a way," Trump said, referring to the agreement between the companies. "And the biggest people in the world signed that, and I signed it as president. And it really is a form of protection."

Trump released the one-page document on his Truth Social platform - which he had earlier described as "morally binding" - and that urged four layers of controls in companies.

First, the document asks companies to partner with an external auditor or evaluator to carry out independent assessments of whether a company's AI control and monitoring systems were operating as intended.

Second, it urges companies to designate an independent committee of the board of directors to oversee and receive reports from internal and external auditors and from teams operating AI controls.

Third, it asks for internal controls to monitor capabilities and alignment of AI models during training and deployment around areas like cybersecurity, biosecurity, and chemical threats to ensure that models "do not hack or access technical systems in unintended ways."

And fourth, it requests that companies have an internal team to ensure "all of the controls, monitoring and detection are operating as intended."

Luncheon spills into public view

What began as a closed-door lunch with technology executives spilled into public view when Trump joined House Speaker Mike Johnson and a lineup of CEOs on the White House driveway, an area rarely used by the president. He took questions for about 30 minutes on AI safety and other topics with the CEOs standing shoulder to shoulder behind him.

The unusual scene came as pressure has mounted on the administration to address the risks and rapid spread of AI, and underscored the alliance Trump has cultivated with leaders of the tech industry. 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.

Zuckerberg, pulled forward from the crowd of billionaire executives by Trump to speak, said that the companies agreed to develop "robust internal controls" for their AI systems, adding that auditors would check those procedures and report back to company boards. Trump said he is considering setting up a 10-person board to police the safety of AI tools, without specifying who would make up that group.

The document was signed by Trump and the bosses of Google, Anthropic, Meta, OpenAI, social media platform X/Twitter, and NVIDIA.

'Makes sense' to transform steps into laws, White House says

"Over time, it may make sense to codify these steps into laws or regulations. Regardless of whether this is required of companies, we believe that implementing these controls and audits is critical to ensuring a safe future for everyone, and each of our companies is committed to doing this," the White House accord said.

"The participating companies will meet regularly to establish standards and best practices to improve the safety of their systems," it added.

Trump doubles down on data centers

Trump separately told reporters at the White House that the tech companies would work to make communities "happy" about the construction of data centers.

Data centers are "going to do things for the community that will be really financially beneficial," he told reporters. "These big, powerful, very rich, very smart companies are going to be making massive contributions to communities."

Trump met with the executives the same day his aides announced a new AI tool that they said will make it easier for the public to use government services.

Silicon Valley executives attended the launch event on Tuesday morning, including Meta's Dina Powell McCormick and Sequoia Capital partner Shaun Maguire. Airbnb co-founder Joe Gebbia led the design of the new tool. Silicon Valley companies Google, SpaceX, Meta and 8VC were listed in the launch event's program as sponsors.

Trump at the launch event said he does not want to work with Chinese President Xi Jinping on governing AI technology, adding that doing so would make it harder for US companies to build better AI products than their Chinese counterparts.

Trump also said that he discussed AI with Xi when he visited the US last week, adding that the US already works with Xi's government "in so many ways."