US AI rules are taking shape as the General Services Administration (GSA) drafted fresh guidelines requiring AI suppliers seeking federal contracts to allow their models to be used for “any lawful” purpose, according to a draft document seen by the Financial Times on March 9, 2026.
The proposed guidelines would require AI vendors to grant the government an irreversible licence to deploy their technology across all legal uses. Moreover, the rules target perceived bias in artificial intelligence outputs, stating contractors must provide “a neutral, non-partisan tool that does not manipulate responses in favour of ideological dogmas such as diversity, equity, inclusion.”
US AI Rules Target Bias and Foreign Compliance
The proposed guidelines also include a clause requiring vendors to disclose whether their models have been modified to comply with foreign regulations, such as the European Union’s Digital Services Act. The GSA plans to consult the industry before finalising the guidelines, the Financial Times reported.
The move reportedly follows an executive order from US President Donald Trump targeting what he described as “woke” AI models. However, the GSA has not yet confirmed a timeline for finalising the rules.
Pentagon Cancels $200M Anthropic Contract
The draft guidelines follow a dispute between the US Department of War (DoW) and AI company Anthropic. The company refused to grant the Pentagon unrestricted rights to deploy its models, citing concerns about domestic surveillance and lethal autonomous weapons.
The Pentagon subsequently cancelled a $200 million contract with Anthropic. DoW Secretary Pete Hegseth wrote on X that “Anthropic delivered a master class in arrogance and betrayal,” adding its “true objective” was “to seize veto power over the operational decisions of the US military.”
“Anthropic delivered a master class in arrogance and betrayal. Its true objective was to seize veto power over the operational decisions of the US military.”
Pete Hegseth, US Department of War Secretary
Following the clash, President Trump issued a directive to cancel Anthropic contracts and implement a six-month phase-out across defence and other federal agencies.
OpenAI Steps In With Pentagon Deal
Rival developer OpenAI moved quickly to fill the gap, announcing a deal to deploy its AI models on the Pentagon’s classified networks. CEO Sam Altman said the agreement would be amended to clarify the technology would not be “intentionally used for domestic surveillance of US persons and nationals.”
Altman stated the agreement includes “red lines” against mass surveillance and autonomous weapons. However, the deal drew immediate scrutiny from within OpenAI itself.
OpenAI Hardware Leader Resigns Over Governance Concerns
OpenAI hardware leader Caitlin Kalinowski announced her resignation in a post on X on Saturday, March 7, 2026. Kalinowski acknowledged that “AI has an important role in national security,” but warned that “surveillance of Americans without judicial oversight and lethal autonomy without human authorisation are lines that deserved more deliberation than they got.”
“Decisions of such scale were too important to be rushed.”
Caitlin Kalinowski, Former OpenAI Hardware Leader
Kalinowski added the deal raised broader cybersecurity and governance concerns. Her departure signals growing internal tension at OpenAI over the company’s expanding role in US defence contracts.
The GSA’s proposed framework, combined with the Pentagon’s rapid pivot from Anthropic to OpenAI, suggests the US government is moving to establish firmer control over how AI technology is deployed across federal agencies. The economy of AI contracting in the defence sector is shifting rapidly, with billions of dollars and significant ethical questions now at stake.