Preventing Woke AI in the Federal Government
The order directs all federal agencies to procure only large language model (LLM) AI systems that comply with two "Unbiased AI Principles" — truth-seeking and ideological neutrality — and bars the purchase of AI tools the order characterizes as embedding DEI ideology into outputs.
It is the first executive order to impose procurement-based content standards on commercial AI products sold to the federal government, requiring vendors to certify their models do not suppress factual information or manipulate outputs along ideological lines.
What this order does
What it orders
The order directs all federal agency heads to procure only LLMs that adhere to two defined Unbiased AI Principles: (1) truth-seeking — models must be truthful, historically accurate, and scientifically objective; and (2) ideological neutrality — models must not encode partisan or ideological judgments, including concepts the order labels as DEI, into outputs. New federal contracts for LLMs must include these requirements and must specify that decommissioning costs are charged to vendors that fail to cure noncompliance. Agencies must also revise existing LLM contracts where practicable.
The policy takes immediate effect as a directive, but the operative contracting requirements do not apply until OMB issues implementing guidance. Exceptions are made for national security systems. The order does not regulate AI in the private marketplace — its reach is limited to federal procurement.
Who it affects
All federal executive agencies purchasing AI language model tools, and commercial AI vendors seeking federal contracts. Vendors must demonstrate their LLMs comply with the Unbiased AI Principles or face contract termination and decommissioning costs. Private-sector AI use outside federal procurement is unaffected.
Why it matters
Federal agencies that rely on commercial AI tools for operations, research, or public services must now evaluate those tools against new content standards before or after procurement. Vendors whose AI models include content policies the order defines as ideological face potential disqualification from federal contracts or termination liability.
What must happen and when
How the order is supposed to work
OMB issues guidance within 120 days, developed in consultation with the procurement, General Services, and technology policy offices. That guidance sets technical compliance standards and permits vendors to satisfy the transparency requirement through system-prompt or specification disclosure rather than revealing model weights. Within 90 days of the guidance, each agency adopts internal compliance procedures. New contracts include cure-period and decommissioning-cost provisions; existing contracts are revised where practicable. National security systems are carved out. No independent enforcement body is created — compliance runs through standard federal contracting mechanisms.
Actions and deadlines
- OMB Director issues guidance to agencies implementing the Unbiased AI Principles procurement requirements
- Agency heads include Unbiased AI Principles compliance terms in all new LLM contracts
- Agency heads revise existing LLM contracts to include Unbiased AI Principles terms where practicable
- Agency heads adopt internal procedures to ensure procured LLMs comply with the Unbiased AI Principles