GSA Looks to Follow DOD’s Lead with New AI Clause for Lawful Government Use

FedScoop
FedScoopMar 26, 2026

Why It Matters

The clause could redefine federal AI procurement, favoring U.S.-based providers and expanding government authority to use AI across all lawful activities, reshaping industry risk and compliance landscapes.

Key Takeaways

  • GSA proposes AI contract clause mirroring DOD’s policy
  • Clause mandates use of only American AI and unbiased standards
  • Government gains right to use AI for any lawful purpose
  • Draft references Anthropic dispute, potentially lifting usage restrictions
  • Comment deadline extended to April 3, affecting industry compliance timeline

Summary

The General Services Administration unveiled a draft amendment to its Multiple Award Schedule that would embed a new AI‑specific clause, echoing the Department of Defense’s recent policy after the Anthropic‑Pentagon dispute.

The language sets four core requirements: government‑wide data use rights, a definition of “unbiased” AI, a mandate that only U.S.-origin AI be procured, and a contractor‑enforced compliance framework. It also grants the government authority to employ the technology for any lawful purpose, effectively removing previous vendor‑imposed usage limits.

The draft directly references the Anthropic controversy, where the company argued that a DOD‑wide “lawful use” clause would have nullified its restrictions on lethal autonomous weapons and mass surveillance. Industry first saw the proposal on March 6, with the comment period now pushed to April 3.

If adopted, the clause could reshape federal AI sourcing, pressuring vendors to relocate development to the United States and tighten bias controls, while raising compliance costs for contractors and influencing broader government AI strategy.

Original Description

The General Services Administration has proposed draft contract language that would define the government’s relationship with AI service providers in a major federal acquisition program in the aftermath of the fallout between the Defense Department and Anthropic. The proposed language includes terms and conditions for government data use, defines what it means for AI to be unbiased, creates a requirement to use only “American AI,” and establishes a responsibility for contractors to enforce terms and conditions on the AI they deploy.
#gsa #federal #government #technology #news #technews #airegulation #trumpadministration #trumpai #governmentnews #federalgovernment #governmentcontracting #acquisition #govtech #govt #usgovernment #artificialintelligence #ai #ainews #aitechnology #anthropic #anthropicai #anthropicclaude #anthropicban #supplychain #supplychainrisk #masssurveillance #leathal #autonomousweapons #autonomousai #autonomouswar #multipleawardschedule #termsandconditions #tandco #americanai #unbiasedai #unbiased #governmentdata #datause #aiservices #claude #claudeai #dod #departmentofdefense #pentagon #petehegseth #donaldtrump #darioamodei #lawsuit #legalchallenges #militarytechnology #defensetechnology #usdefense #usmilitary #miltech #militarynews #defensenews #aigovernance #governance
Follow FedScoop on Social Media
About FedScoop
FedScoop is the leading tech media brand in the federal government market. FedScoop gathers top leaders from the White House, federal agencies, academia and the tech industry to discuss ways technology can improve government and identify ways to achieve common goals. https://fedscoop.com/

Comments

Want to join the conversation?

Loading comments...