On a crisp Tuesday, President Donald Trump called a one‑day meeting of America’s top AI bosses to the White House. The heads of major firms – from OpenAI to Meta, Nvidia to SpaceX – sat around a gleaming table and signed a pact that the president described as a “morally binding” deal to keep their systems safe.
The agreement, signed by CEO Sundar Pichai, Dario Amodei, Mark Zuckerberg, Greg Brockman, Elon Musk and Jensen Huang, sets out a series of internal and external checks. Companies must build safeguards so models run as intended, flag any problems, and work with independent auditors to ensure their platforms cannot “hack or access technical systems in unintended ways.” Trump called the accords a “constitution” for AI.
He followed up the summit with an executive order that tells all U.S. agencies to drop the use of the phrase “artificial intelligence” in favour of Super Intelligence (SI). “These advances are no longer just artificial intelligence – they are a new era of Super Intelligence,” the order reads.
Trump also pledged to expand data‑centres – the power plants that run AI – and to step in as a board that can oversee their safety. The call for more data‑centres came amid a poll that found 65% of voters opposed new facilities in their neighbourhoods. “These companies will bring big jobs and prosperity,” he said, urging that they serve the communities they build in.
The summit ran a mixed reception. Industry insiders praised the “historic conversation,” while academic experts label the pact “deeply unimpressive” because it leans on companies to define safety thresholds and offers no clear penalties for breaches. Kim Weatherall, a law professor, called it a distraction that should be ignored.
In a UN address, Trump decried the word “artificial” as making the technology feel fake – a phrase some tech leaders are already adopting. However, analysts warn the new term is unlikely to catch on outside the U.S. and may only mark a political re‑branding rather than a real shift in how AI is regulated.
