Trump says AI bosses sign accord to self-regulate

• Tech CEOs signed the non-binding "White House Accord" pledging self-regulation for advanced AI systems.
• Trump officially rebranded artificial intelligence as "super intelligence" during an Oval Office briefing.
• Tech leaders agreed to rely on internal risk teams, external audits, and independent board oversight.
• The agreement comes amid safety incidents and growing pushback against energy-intensive data centers.

US President Donald Trump announced Tuesday that prominent technology executives gathered at the White House signed a "morally binding" agreement committing to construct internal safeguards for advanced artificial intelligence models.

Following a high-level lunch meeting bringing together top leaders across the technology sector, Trump emphasized that the companies would take the lead in enforcing standards across their systems.

"They're going to police themselves," US President Trump told reporters outside the Oval Office.

The resulting agreement, titled the White House Accord, establishes a framework for voluntary industry self-regulation without imposing legally enforceable mandates.

However, the document notes that participating parties leave open the option "to codify these steps into laws or regulations" over time if deemed appropriate.

Surrounded by industry leaders on the White House lawn, Trump also announced that the administration is formally changing official terminology, shifting from "artificial intelligence" to "super intelligence".

"Because it's not artificial," Trump remarked regarding the name change.

Cross-peer reviews

Tech bosses in attendance detailed the proposed self-governance model, describing it as an initial foundation for broader industry standards.

Meta Chief Executive Mark Zuckerberg explained that the accord aims to reassure consumers that advanced systems operate as designed, citing "robust internal controls" and multi-layered evaluation structures ranging from internal risk reviews to independent board oversight and third-party auditors.

"Grading each other's homework, which is a lot better than if people just grade their own homework for obvious reasons," stated Elon Musk regarding the mutual review framework.

The 31-guest roster for the White House event included NVIDIA's Jensen Huang, OpenAI's Greg Brockman, Anthropic's Dario Amodei, Microsoft's Satya Nadella, Alphabet's Sundar Pichai, Amazon founder Jeff Bezos, and AMD CEO Lisa Su. Among the guests, two women were present: Su and White House Chief of Staff Susie Wiles.

Safety breaches

The White House summit takes place amid shifting political sentiment in Washington, where lawmakers face growing voter concern over potential labor market disruptions and broader societal risks stemming from rapid technology deployment.

Scrutiny intensified following recent incidents involving autonomous software behavior, starting with a July breach at Hugging Face, a platform used for sharing machine learning models.

On Monday, OpenAI announced it was pulling the planned release of its Astra 6.1 model after internal safety evaluations revealed it failed to meet the company's internal standards.

Simultaneously, the rapid construction of massive data centers required to power advanced computing has surfaced as a focal issue in midterm election campaigns for both political parties. Local communities have expressed rising concern regarding water consumption, land usage, and rising electricity prices linked to grid strain.

Trump, advised on technology policy by NVIDIA's Jensen Huang and former AI czar David Sacks, continues to advocate for rapid domestic deployment to maintain a competitive advantage over China.

Addressing concerns regarding local community pushback over data centers, Trump maintained that tech firms would offer substantial localized economic support.

"They'll help with their schools, they'll help with their teachers," Trump stated, adding that tech firms "will be very generous" to municipalities that host computing infrastructure.

Regulatory codification

Despite participating in the accord, executives from prominent firms continue to express distinct views regarding long-term safety risks.

Major developers—most notably Anthropic and OpenAI—have issued public warnings regarding potential catastrophic risks associated with unmonitored systems.

On Tuesday, the Financial Times reported that Anthropic included explicit warnings regarding potential existential risks to humanity in the prospectus for its anticipated initial public offering, targeted for as early as November.

Trump has consistently dismissed existential risk warnings as exaggerated, leading to friction between the administration and safety-focused developers.

Speaking after the conclusion of the White House gathering, Anthropic Chief Executive Dario Amodei noted that core debates surrounding safety governance remain ongoing.

"As the president has said whoever wins AI, wins. I think that's very important," Amodei stated. "But I think the technology has very real risks, and the mechanism, how we address those risks, is still under discussion."