President Donald Trump and House Speaker Mike Johnson announced on Tuesday that several major artificial intelligence companies have signed a voluntary agreement establishing standards intended to increase industry oversight. The initiative aims to have companies commit to more transparent and accountable reviews of their own technologies.
Johnson told reporters the pact is "a statement of principles ... a statement of standards, commitments that are voluntary on behalf of the industry. It sets forth expectations." He added that oversight and transparency are needed and that the industry leaders present would work toward those goals.
During the announcement, Trump also proposed renaming the technology from “AI” to “SI,” meaning superintelligence.
What the companies say they will do
Mark Zuckerberg outlined measures that participating firms expect to implement, including internal risk reviews and external audits. He said company leaders drafted a list of principles covering reviews of internal controls and mechanisms to detect problems with the technology, potentially involving multiple layers of auditing.
The Information reported that top AI labs Anthropic, OpenAI and Google had been coordinating before the White House meeting on a self-regulation framework intended to align testing and auditing of the technology.
Concerns and loose ends
Sources working in AI told Axios they worry that self-policing will not resolve the safety problems currently confronting the industry. The scope of the agreement and how it would be enforced remain unclear because the full terms have not been released.
When asked for comment Tuesday, the White House replied to Axios: "Turn on TRUTH notifications," referencing the president’s Truth Social account. Trump described the accord as "morally binding," adding, "If something goes wrong, their companies are at stake ... They're really going to be policing each other."
Atmosphere at the event and next steps
Anthropic CEO Dario Amodei addressed reporters and emphasized the safety risks of the technology; Trump, Zuckerberg and former AI czar David Sacks were seen laughing during that exchange.
At a panel following the event, Nvidia CEO Jensen Huang adopted the term "SI" for the technology — the label Trump has been promoting — after briefly calling it AI. "I am going to get good at this," Huang said.
Further details on how the voluntary standards will function are expected after Tuesday's meeting between AI executives and Washington officials. An AI czar is also expected to be selected within the next three to four days, and Trump intends to consult with Speaker Mike Johnson and Senate Majority Leader John Thune about that choice.
