Trump and Tech Titans Sign AI Self-Policing Pact, Ignoring Federal Oversight

The White House got crowded fast. President Donald Trump sat down with the biggest names in tech and walked out with a signed voluntary agreement that puts AI oversight squarely on the companies building it.

The accord, called the “Joint Commitment on Frontier Responsibilities,” was announced at the White House and leans hard on developers to manage their own risks. No mandatory rules. No federal enforcement agency calling the shots. Just internal controls, independent external audits, and a vague promise that board-level oversight will keep AI models from doing things nobody asked them to do. The agreement leaves the door open for these voluntary commitments to eventually become formal laws — but that’s not on the table yet. For now, it’s self-policing or nothing.

Trump was blunt about it.

“They understand that they have to self-police,” he said. And he meant it literally. The accord pushes companies to build accountability into their own structures rather than wait for Washington to hand them a rulebook. Trump’s position on China made that even clearer — he shut down any talk of a joint US-China AI venture during the meeting, full stop. Staying ahead of Beijing matters more to him right now than building international frameworks.

Who Signed and What They Agreed To

The names on the accord read like a who’s-who of the industry. Google CEO Sundar Pichai signed. So did Anthropic CEO Dario Amodei, Meta CEO Mark Zuckerberg, OpenAI President Greg Brockman, Nvidia CEO Jensen Huang, and xAI CEO Elon Musk. Pretty much every major player with a frontier AI model in production showed up.

The meeting covered a lot of ground — AI safety, industry growth, data center development. That last one matters more than it sounds. Data centers are the physical backbone of everything these companies are scaling, and the infrastructure demands are enormous. Building faster AI means building more of them, and that raises real logistical and energy questions that didn’t get answered publicly.

The accord’s core ask is straightforward: don’t let your AI systems go rogue. Board-level oversight is specifically called out as a requirement to stop AI models from accessing technical systems in ways nobody intended. It’s a high-level commitment, and the details of how each company actually implements it are, for now, left to each company.

More context: SEC Charges Foreign Fraudsters for $15 Million WhatsApp Crypto Scam Targeting U.S. Investors

Not exactly airtight. But it’s a start.

The “Super Intelligence” Order

Same day, Trump signed a separate executive order. It directs the entire executive branch to stop using the term “Artificial Intelligence” and replace it with “Super Intelligence.” The order’s reasoning is that the new term better captures what these technologies are actually becoming — their promise, their speed, their scope.

It’s a terminology shift, but terminology matters in policy. How a government names something shapes how it regulates it, funds it, and talks about it publicly. Calling it “Super Intelligence” is a signal that the administration sees this technology as something bigger than the current framing allows. Whether that framing eventually drives different legislation is unclear.

The accord itself seems to anticipate that question. Its flexible structure was built to allow voluntary commitments to harden into binding regulations if self-regulation turns out to be insufficient. That’s probably the most interesting clause in the whole document — an admission, built right into the agreement, that the industry might not be able to handle this alone.

For crypto markets and blockchain developers, the governance dynamics here aren’t abstract. The same debate over self-regulation versus formal oversight has played out in crypto for years. Voluntary commitments, industry-led standards, promises of internal audits — crypto went through all of it before regulators stepped in hard. AI seems to be starting from a similar place.

See also: Nvidia Unveils AI Safety Platform to Contain Rogue Agents After Breach Incidents

And the tech giants signing this accord aren’t strangers to crypto infrastructure either. Nvidia chips power a huge share of both AI training and crypto mining operations. Google, Meta, and others have explored blockchain applications at various points. How AI governance frameworks develop will almost certainly bleed into how digital asset oversight gets shaped over the next few years.

The accord’s board-level oversight requirement is worth watching specifically. If companies actually enforce it — and that’s a big if — it creates a precedent for holding executives personally accountable when AI systems misbehave. That kind of accountability structure is something crypto regulation has pushed for repeatedly and rarely gotten.

No timeline was given for when or whether the voluntary commitments might become law.