Anthropic and OpenAI face Trump admin pressure on AI safety and China competition
In brief
- Trump administration imposed export controls on Anthropic models and curtailed OpenAI's GPT-5.6 rollout
- Over 1,100 AI lab employees signed petition for deliberate slowdown in AI advancement
- Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman backed careful pacing of development
- Anthropic filed IPO paperwork at $965 billion valuation amid regulatory tension
The regulatory squeeze
In June 2026, the Trump administration directed the US Commerce Department to impose export controls and access restrictions on Anthropic's Mythos 5 and Fable 5 models, citing cybersecurity concerns. That same month, OpenAI's GPT-5.6 rollout was curtailed at the White House's request, limited to vetted partners rather than a broad public release.
The moves marked a sharp pivot. After years of deregulation rhetoric, the government pivoted to imposing stricter checks on AI capabilities through executive orders. Additionally, foreign nationals saw temporary access to certain Anthropic models disabled entirely.
The internal revolt
Yet the pressure isn't only top-down. Over 1,100 employees from various leading AI labs signed a 'Pacing the Frontier' petition in early 2026, advocating for a deliberate slowdown in AI advancement. The internal anxiety found a public voice when Anthropic CEO Dario Amodei published an essay in September 2026 arguing for pacing AI capabilities more carefully, centered on risks like recursive self-improvement and rogue AI instances.
Remarkably, OpenAI CEO Sam Altman publicly supported Amodei's position on pacing AI development. Both executives went further, calling for independent evaluators to assess model safety before deployment.
The geopolitical counterpressure
President Trump sees no room for caution. "Whoever wins AI wins," Trump stated, framing the technology race with China as an existential competition that America must lead. The Trump administration insisted the US cannot afford to slow down.
This puts Anthropic and OpenAI in an impossible position. Anthropic filed IPO paperwork in mid-2026 with a valuation reaching $965 billion, but the company now operates under export controls and access restrictions. Neither firm can claim to investors that the path forward is clear. They're being told simultaneously to move faster and slower—to innovate at scale while proving they've thought through existential risks, all while geopolitical pressure mounts to keep pace with Beijing.
