US government gets early access to AI models for security testing
The US government has secured early access to new AI models from Microsoft, Google, and xAI for national security testing through the Center for AI Standards and Innovation. The agreement aims to evaluate the models' capabilities and security risks before they are widely deployed.
The US government's Center for AI Standards and Innovation has agreed to receive early access to new AI models from Microsoft, Google, and xAI for national security testing. Microsoft will work with US government scientists to test AI systems and develop shared datasets and workflows. The agreement fulfills a pledge made in July 2025 to partner with tech companies to vet their AI models for national security risks. The US government is growing concerned over the national security risks posed by powerful AI systems, including cyberattacks and military misuse. The Center has already completed over 40 evaluations, including on cutting-edge models not yet available to the public. Developers provide versions of their models with safety guardrails stripped back to allow the center to probe for national security risks.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.