WorkThe story, in brief

White House reverses course and restores UK AI Safety Institute's access to frontier models

US-UK AI safety data-sharing just became diplomatic. The White House reversed its ban on frontier model access to Britain's AI Safety Institute after pressure from London's AI Minister.

Illustration of a transparent lens revealing connected networks across layers of paper.
Exploring the next frontier of AI research.AI illustration by KeyNews
The KeyNews take

Why it matters

A reversal of Trump administration policy on AI model access to allied nations signals shifting US posture on frontier model sharing and international AI governance coordination. Practitioners should track whether this affects their own cross-border AI safety and evaluation workflows.

The key facts

9 to know
  1. White House reversed September 2026 ban on UK AI Safety Institute access to frontier models

  2. Ban originally required Anthropic, Google DeepMind, and OpenAI to halt model-sharing with British agency

  3. Reversal driven by diplomatic pressure from UK AI Minister Kanishka Narayan

  4. Affects early access to frontier models from three major labs

  5. White House reversed September 2026 ban on UK AI Safety Institute's early access to frontier models

  6. Ban covered Anthropic, Google DeepMind, and OpenAI models

  7. Reversal followed diplomatic pressure from UK AI Minister Kanishka Narayan

  8. Original directive came from Trump administration

  9. UK AI Safety Institute regains early-access model evaluation capability

The story so far

Earlier coverage of this storyline

  1. Trump unveils his new Super Intelligence ForceTechCrunch AI
  2. AI doesn’t have to wipe out humanity for the governance model to failCIO
  3. Microsoft's Nadella says AI needs an ‘emergency brake’ that humans controlCNBC Technology
  4. Satya Nadella says we should assume all AI models are ‘compromised’The Verge AI
  5. AI’s quiet safety gatekeepers are stepping into the spotlightCNBC Technology
  6. This story

Go to the source

The Decoderthe-decoder.com

Publisher excerpt: The White House has reversed course and restored the UK's AI Security Institute's early access to AI models from Anthropic, Google DeepMind, and OpenAI. In September, the Trump administration had told the companies to stop sharing their models with the British agency. Diplomatic pressure from AI…
Read original report
Back to today's editionMore work news

Keep reading

Related stories

More from Work