Technology & OSINT — 2026-03-23

Trump Admin AI Policy Sends Conflicting Signals: Anthropic Banned While GSA Mandates 'American AI' in Contracts

The Trump administration's push for federal AI adoption is undermined by contradictory actions. While the White House released a vendor-friendly AI Action Plan last July, the Pentagon designated Anthropic's Claude a supply chain risk and ordered all agencies to stop using it, with defense contractors required to certify non-use. Simultaneously, GSA released a nine-page draft clause for schedule contracts requiring vendors to use only 'American AI systems,' disclose all AI in service delivery, and grant government ownership of data and custom developments. George Washington University's Jessica Tillipman observed both actions 'send a really bad message to industry.' Experts warn the restrictions may push vendors and agencies away from GSA schedules toward alternative procurement vehicles. The GSA comment period was extended to April 3.

Analysis
The prior IC brief reported Maven's designation as a permanent program of record. The GSA clause creates a parallel procurement chokepoint: IC agencies using GSA schedules must now certify their AI tools are 'American,' a term with no statutory definition, while simultaneously being barred from the American company (Anthropic) whose model underpins their primary targeting system.

1 sources

Trump administration clouds up its push for AI in government - Federal News Network