The U.S. government has drafted stringent rules for artificial-intelligence contracts, demanding companies grant "any lawful" use of their AI models. This move comes amid a conflict between the Pentagon and AI firm Anthropic. The Defense Department previously flagged Anthropic as a "supply-chain risk," blocking its technology for military work due to disagreements over safeguards.
A draft of the guidelines indicates that AI companies seeking government business must provide the U.S. with an irrevocable license for all legal purposes. This initiative from the General Services Administration aims to bolster AI procurement across civilian agencies and mirrors potential Pentagon actions for military contracts.
An official stated it would be "irresponsible and dangerous" to maintain a business relationship with Anthropic. Consequently, the GSA has terminated Anthropic's OneGov deal, removing access for federal branches through pre-negotiated contracts.
The proposed GSA rules also require AI contractors to avoid embedding partisan or ideological judgments into AI outputs and to disclose any modifications made to comply with non-U.S. regulatory frameworks.