Who's who · AI security & safety
Center for Advancing Innovation and Standards for Super Intelligence
US government centre at NIST that tests commercial AI models; formerly CAISI and the US AI Safety Institute
At a glance
- Founded
- November 2023
- Based in
- Gaithersburg, United States
- People
- Arvind Raman, Acting director; Director of NIST
- Known for
- Frontier model evaluations, Assessments of Chinese models, Developer agreements
- Funding
- US federal government (NIST, Department of Commerce)
- Parent
- NIST, US Department of Commerce
- Former names
- US AI Safety Institute; Center for AI Standards and Innovation (CAISI)
- Website
- https://www.nist.gov/caissi
The Center for Advancing Innovation and Standards for Super Intelligence (CAISSI) is the US government's main point of contact with industry for testing and joint research on commercial AI systems. It sits within the National Institute of Standards and Technology (NIST), part of the Department of Commerce.
It began as the US AI Safety Institute, announced in November 2023. In June 2025 it was rebranded the Center for AI Standards and Innovation (CAISI). It took its current name at the end of September 2026, after a 29 September executive order told federal agencies to say "Super Intelligence" in place of "artificial intelligence".
The centre makes voluntary agreements with AI developers, evaluates models for national security risks in cybersecurity, biosecurity and chemical weapons, and publishes voluntary guidelines. Its 2026 work includes assessments of DeepSeek V4 Pro and Z.ai's GLM-5.2 and GLM-5.3, and a joint study of Kimi K3 with the UK AI Security Institute. NIST Director Arvind Raman has been acting director since Chris Fall left in July 2026.
Sources
- NIST, Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), 2026-10-02
- The White House, Fact Sheet: President Donald J. Trump Inaugurates The Era of Super Intelligence, 2026-09-29
- Nextgov/FCW, Commerce rebrands its AI Safety Institute, 2025-06-04
- Nextgov/FCW, NIST AI safety center lead departs, 2026-07-21
Facts last checked 2026-10-02