Microsoft, Google, xAI give US access to AI models for security testing | Business and Economy News
Microsoft, Google, xAI give US access to AI models for security testing | Business and Economy News
Publish Date: 2026-05-05 13:04:00
Source Domain: www.aljazeera.com
- The U.S. federal government, through the Center for AI Standards and Innovation (CAISI), has reached an agreement with major tech companies including Microsoft, Google, and xAI to allow testing of their new AI models for national security purposes before they are deployed.
- The objective of this agreement is to identify and mitigate potential security risks associated with advanced AI systems, ranging from cyberattacks to military misuse.
- This initiative fulfills a previous pledge by the Trump administration to partner with the tech industry to evaluate AI models for national security risks.
- Microsoft has committed to working with U.S. government scientists to develop shared datasets and workflows for testing AI systems to uncover unexpected behaviors.
- This move builds on previous agreements under President Biden’s administration, which aimed to develop AI tests, definitions, and voluntary safety standards.
- The Pentagon also recently announced an agreement with seven major tech companies to use AI systems across classified networks to improve decision-making for warfighters.
- Notably, Anthropic was not included in these agreements following a dispute with the Trump administration over the ethics and safety of AI usage in war.
