# White House Restricts UK AI Institute Access

The White House has directed leading US AI developers OpenAI and Anthropic to withhold pre-release access to advanced AI models from the United Kingdom's AI Security Institute (AISI) pending completion of a domestic review by the US Center for AI Standards and Innovation (CAISI). CAISI currently operates without a permanent director and lacks the comprehensive evaluation infrastructure that AISI possesses, potentially impacting the robustness and independence of AI safety assessments available to government and enterprise buyers. This decision reflects ongoing US efforts to centralize AI safety oversight domestically and may influence procurement strategies involving AI model validation and compliance.

- **Why this matters:** Procurement professionals should note that US federal agencies and contractors may prioritize AI safety certifications and evaluations conducted or endorsed by CAISI, affecting vendor eligibility and compliance requirements.
- The absence of a permanent CAISI director and limited infrastructure may delay or constrain AI safety validation processes, impacting acquisition timelines and risk assessments.
- Companies developing or supplying AI technologies should prepare for evolving US standards and potential restrictions on international collaboration or data sharing with entities like AISI.
- Organizations involved in AI procurement should evaluate the implications of this policy on supply chain security, model certification, and cross-border technology partnerships.

**Jurisdictions:** federal
**Industries:** Information Technology
**Topics:** Artificial Intelligence
**Published:** September 28, 2026

### Government Entities
- Office of the National Cyber Director (ONCD)
- Center for AI Standards and Innovation (CAISI)
- AI Security Institute (AISI)
- National Institute of Standards and Technology (NIST)
- United Kingdom Government

### Vendors
- OpenAI (AI model developer)
- Anthropic (AI model developer)

### Key Quotes
> Shared global principles and standards are needed, and the UK works hand in glove with the US and leading AI labs.
> — Andy Burnham, Prime Minister, United Kingdom

### Sources
- [White House Bars UK AI Safety Institute From Frontier AI Testing; CAISI, US Body, Has No Director](https://www.techtimes.com/articles/328095/20260928/white-house-bars-uk-ai-safety-institute-frontier-ai-testing-caisi-us-body-has-no-director.htm) - techtimes.com