Musk, Zuckerberg, Huang lobbied Trump to stall AI rules after 14 July proposal

The pushback did not come out of nowhere. According to the report, it began after Google DeepMind's Demis Hassabis put forward a proposal on 14 July advocating a new regulatory body to set standards for AI safety — a framework supported by Anthropic's Dario Amodei. Musk, Zuckerberg and Huang then reportedly took their concerns straight to the Oval Office, making this a case of industry leaders lobbying the US President directly rather than working through a public consultation.

The public positions that followed differ in kind. Musk, at the All-In Summit in Los Angeles on Monday, argued that leading AI companies should test each other's frontier models before release, so that competitors — not the developers themselves — raise the alarm. Nvidia's Huang told Salesforce's Dreamforce conference in San Francisco on Tuesday that existing laws are sufficient and that builders who lack confidence in a product's safety should simply not release it. Trump's own posts on Truth Social earlier this week described AI fears as a "hoax" carried out by "traitors" and warned that any slowdown could give China the upper hand.

What happens next appears to hinge on whether self-testing among competitors gains traction as a substitute for a formal standards body, and on how the split between the companies plays out. Anthropic's Amodei and OpenAI CEO Sam Altman have warned of devastating consequences without proper oversight, so the industry is not speaking with one voice here — a division that may shape how much appetite remains in Washington for new rules.