U.S. to exempt open-weight AI models from new safety testing, report says

A report from Neowin indicates that the U.S. government intends to exclude open-weight AI models from new safety testing protocols. The move is seen as an effort to avoid hindering innovation in the rapidly evolving AI sector, where open-weight models are widely used by developers and researchers. The policy, if implemented, would apply to models with public weights and would not subject them to the same rigorous evaluation as closed proprietary models. The decision comes amid ongoing debates about the risks and benefits of open-weight AI, with proponents arguing that it fosters transparency and collaboration, while critics worry about potential misuse.