Center for AI Standards and Innovation
In plain wordsThe US government's AI testing and standards center, part of the National Institute of Standards and Technology. It evaluates advanced AI models, American and foreign, for national-security risks and works with companies on voluntary testing.
A center of the National Institute of Standards and Technology (NIST) that is the federal government's designated partner for AI companies on model testing and research, with a remit centered on national security. Its work turns on evaluations: voluntary testing under agreements with developers, unclassified assessments of capabilities with security implications (chiefly cyber, biological and chemical), and examinations of rival states' systems for backdoors and foreign influence, as in its published evaluations of Chinese open-weight modelsopen-weight model, n.: An AI model whose trained internal numbers are published for anyone to download and run on their own computers, even if the data and code used to make it stay private. from DeepSeek, Moonshot AI and Z.ai. Beyond testing, it drafts guidance on measuring and strengthening the security of AI systems and speaks for the United States in international standards work, where it is charged with resisting what the administration calls burdensome foreign regulation. It also runs an AI Agent Standards Initiative, and in June 2026 it reportedly tested the improved safeguards that preceded the Commerce Department's withdrawal of an export control directiveexport control directive, n.: A US government order, used against AI models in June 2026, telling a company to stop foreigners from using some of its AI models for national-security reasons. against two Anthropic models. After Executive Order 14434Executive Order 14434, prop. n.: A September 2026 order by US President Donald Trump telling federal agencies to say 'super intelligence' (SI) instead of 'artificial intelligence' (AI) in official documents and communications. (September 2026) told agencies to say 'Super Intelligence' instead of 'AI', the heading of NIST's web page for the center read 'Center for Advancing Innovation and Standards for Super Intelligence (CAISSI)', but as of October 2026 no formal renaming had been announced and NIST's own text still used CAISI.
CAISI's evaluation compared the Chinese open-weight model with US frontier models on cyber tasks.
Origin
Created as the US AI Safety Institute in November 2023 under Executive Order 14110Executive Order 14110, prop. n.: A sweeping 2023 order by US President Joe Biden that told federal agencies how to manage AI risks and required makers of the largest AI models to report their safety tests to the government. It was revoked in 2025. and led initially by Elizabeth Kelly; Commerce Secretary Howard Lutnick announced its reorganization and renaming as CAISI on June 3, 2025, dropping 'safety' from its name. America's AI Action PlanAmerica's AI Action Plan, prop. n.: The Trump administration's 2025 strategy for winning a global AI race: cutting regulation, building data centers and power plants quickly, and exporting American AI technology to allies. (July 2025) assigned it further evaluation duties.
Usage note
Its shift from safety toward security and competitiveness paralleled the UK institute's 2025 renaming as the AI Security Institute. It was created by departmental decision rather than by statute; the Future of AI Innovation Act of 2024, which would have codified its predecessor, did not pass.
Related
- AI Safety InstituteAI Safety Institute, n.: A government agency that tests the most advanced AI models for dangerous abilities and studies how to make them safe. Britain and the United States set up the first ones in 2023.
- third-party AI evaluationthird-party AI evaluation, n.: Testing of an AI system by someone other than the company that made it — a government institute, a nonprofit lab or independent researchers — to check what it can do and how risky it is.
- pre-deployment testingpre-deployment testing, n.: Checking a new AI model for dangerous abilities and bad behavior before it is released to the public, sometimes with outside experts or government testers given early access.
- Executive Order 14110Executive Order 14110, prop. n.: A sweeping 2023 order by US President Joe Biden that told federal agencies how to manage AI risks and required makers of the largest AI models to report their safety tests to the government. It was revoked in 2025.
- America's AI Action PlanAmerica's AI Action Plan, prop. n.: The Trump administration's 2025 strategy for winning a global AI race: cutting regulation, building data centers and power plants quickly, and exporting American AI technology to allies.
- Executive Order 14434Executive Order 14434, prop. n.: A September 2026 order by US President Donald Trump telling federal agencies to say 'super intelligence' (SI) instead of 'artificial intelligence' (AI) in official documents and communications.
- NIST AI Risk Management FrameworkNIST AI Risk Management Framework, prop. n.: A free, voluntary guide from the US government's standards agency that helps organizations spot, measure and reduce the risks of the AI systems they build or use.
Sources
- National Institute of Standards and Technology. Center for AI Standards and Innovation (CAISI) web page (headed 'CAISSI' as of October 2026). nist.gov/caisi
Mentioned in
AI Safety InstituteAI Safety Institute, n.: A government agency that tests the most advanced AI models for dangerous abilities and studies how to make them safe. Britain and the United States set up the first ones in 2023., America's AI Action PlanAmerica's AI Action Plan, prop. n.: The Trump administration's 2025 strategy for winning a global AI race: cutting regulation, building data centers and power plants quickly, and exporting American AI technology to allies., Center for AI SafetyCenter for AI Safety, prop. n.: A San Francisco nonprofit that studies how advanced AI could lead to disaster and how to prevent it. In 2023 it gathered signatures for a one-sentence warning that AI could threaten human extinction., Executive Order 14110Executive Order 14110, prop. n.: A sweeping 2023 order by US President Joe Biden that told federal agencies how to manage AI risks and required makers of the largest AI models to report their safety tests to the government. It was revoked in 2025., Executive Order 14434Executive Order 14434, prop. n.: A September 2026 order by US President Donald Trump telling federal agencies to say 'super intelligence' (SI) instead of 'artificial intelligence' (AI) in official documents and communications., NIST AI Risk Management FrameworkNIST AI Risk Management Framework, prop. n.: A free, voluntary guide from the US government's standards agency that helps organizations spot, measure and reduce the risks of the AI systems they build or use., Paul ChristianoPaul Christiano, prop. n.: An American AI-alignment researcher who led the work that made training AI from human preferences practical, the basis of how chatbots are taught to be helpful., pre-deployment testingpre-deployment testing, n.: Checking a new AI model for dangerous abilities and bad behavior before it is released to the public, sometimes with outside experts or government testers given early access., third-party AI evaluationthird-party AI evaluation, n.: Testing of an AI system by someone other than the company that made it — a government institute, a nonprofit lab or independent researchers — to check what it can do and how risky it is., AgentDojoAgentDojo, prop. n.: A testing environment that measures how easily AI assistants that use email, calendars or banking tools can be hijacked by hidden instructions planted in the content they read., Alignment Research CenterAlignment Research Center, prop. n.: A small Berkeley nonprofit started by Paul Christiano that searches for mathematical ways to keep future AI systems, even ones far smarter than people, honest and under human control., export control directiveexport control directive, n.: A US government order, used against AI models in June 2026, telling a company to stop foreigners from using some of its AI models for national-security reasons.