US Halts Security Testing on Open-Weight AI Models

Trump administration signals shift in AI regulation by ending security assessments for open-weight models, allowing unrestricted deployment without government oversight.

Woofun AI reports that the Trump administration notified leading artificial intelligence companies on Tuesday that it will cease conducting security testing on open-weight AI models. This policy shift alters the U.S. government's regulatory stance, as open-weight models permit developers to download, modify, and deploy parameters without closed API restrictions. Previously, entities such as the American Institute for AI Safety (AISI) performed security assessments on advanced models. The White House and relevant departments have not issued public comments regarding this development.