The Office of Management and Budget has released a memorandum outlining contractual requirements to ensure that large language models, or LLMs, procured by the federal government comply with unbiased artificial intelligence principles.
Join us for the Potomac Officers Club’s 2026 Artificial Intelligence Summit on March 18. Book your spot now to explore cutting-edge use cases and be part of the conversation shaping the future of AI in government.
In the seven-page memo dated Thursday, OMB Director Russell Vought wrote that the executive order signed in July identifies two unbiased AI principles. According to the EO, LLMs must be “truth-seeking” by prioritizing historical accuracy, objectivity and scientific inquiry and have “ideological neutrality” by not integrating partisan or ideological judgments into a model’s outputs.
What Are OMB’s Contractual Requirements for Agencies Procuring LLMs?
According to the memo, agencies must secure sufficient information from the vendor to determine whether an LLM complies with unbiased AI principles.
OMB noted that agencies should avoid requirements that force a vendor to divulge specific model weights and other sensitive technical data.
In LLM solicitations, OMB directs agencies to request information from a vendor, including acceptable use policy; information about the model, system or data cards; end user resources; and mechanism for end user feedback.
For enhanced LLM transparency, agencies may seek additional information regarding pre-training and post-training activities, model evaluations, enterprise-level controls and third-party modifications.
According to OMB, agencies must update their procedures and policies by March 11 to ensure that LLM procurement contracts include contractual requirements addressing compliance with unbiased AI principles.
How Does the New OMB Memo Build on the AI Action Plan?
The latest OMB memo delivers on the AI Action Plan released by the White House in July. The plan outlines over 90 federal policy actions across three pillars: accelerating AI innovation; building American AI infrastructure; and leading in international AI diplomacy and security. One of the recommended policy actions cited in the document is to update federal procurement guidelines to ensure that agencies only deal with frontier LLM developers that ensure that their offerings “are objective and free from top-down ideological bias.”
