Artificial Intelligence in Government Procurement

AI is transforming federal procurement from two directions simultaneously: agencies are buying AI systems to improve government operations, and contractors are deploying AI tools to improve how they identify, pursue, and execute government contracts. Both dimensions carry distinct regulatory and risk considerations.

What is Artificial Intelligence in Government Procurement?

AI as a product being procured: Federal agencies are acquiring AI systems for applications including fraud detection, natural language processing for document review, predictive maintenance, cybersecurity threat detection, and decision support for benefits adjudication. Key policy frameworks governing AI procurement include:

• Executive Order on AI (2023): Required agencies to assess AI risks, establish safety standards, and develop acquisition guidance for AI systems, particularly those used in high-impact decision-making.
• OMB Memorandum M-24-10: Directed agencies to implement AI governance frameworks, designate Chief AI Officers, and apply rights-protecting and safety-enhancing practices when procuring AI.
• NIST AI Risk Management Framework: Provides a voluntary framework increasingly referenced in AI procurement requirements to assess and manage AI risks including bias, transparency, and explainability.

AI in the procurement process: Contractors and government procurement offices use AI tools to improve the efficiency and quality of acquisition activities including market research, solicitation analysis, proposal development, evaluation support, and contract management. The use of AI in generating proposal content has raised questions about authenticity certifications, and some solicitations now include representations regarding AI-generated content.

Why AI in government procurement matters for contractors

Contractors who supply AI products to the government must understand the emerging compliance framework around AI safety, bias testing, explainability, and data governance. Contractors who use AI tools in proposal development must understand the applicable representations and certifications, particularly for solicitations that ask about AI-generated content or require human review of AI outputs.

Example

A health IT company develops a machine learning model that predicts hospital readmission risk for the VA. The solicitation requires that the model be documented using the NIST AI RMF framework, that the company provide regular bias audits comparing performance across demographic groups, and that all model predictions include an explainability report identifying the top factors driving each prediction. The company builds these requirements into its solution design and proposal, demonstrating compliance with the agency's AI governance standards.

Frequently Asked Questions

Does the government have a standard contract clause for AI procurement?

As of mid-2026, there is no single standard FAR clause for AI procurement. Agencies are developing their own clause language, and the FAR Council has ongoing rulemaking on AI-related acquisition policies. Contractors should expect significant variation in AI-specific requirements across agencies and solicitations.

Can contractors use AI to write proposals?

There is no blanket prohibition on using AI tools in proposal development. However, solicitations increasingly include representations that the offeror attests to the accuracy and authenticity of their proposals. Using AI to generate factual claims that are not verified creates False Claims Act risk. Best practice is to use AI for drafting and organization while ensuring subject matter experts review and attest to all substantive content.

What is a high-impact AI system under OMB guidance?

High-impact AI systems are those used in contexts where decisions significantly affect individuals' rights, opportunities, or safety, including benefits determinations, criminal justice applications, and healthcare decisions. OMB M-24-10 requires agencies to apply additional safeguards when procuring high-impact AI systems.

How does CMMC interact with AI systems handling controlled information?

AI systems that process Controlled Unclassified Information (CUI) are subject to CMMC requirements just like any other IT system handling CUI. Contractors must ensure their AI platforms, training data environments, and inference systems meet the applicable CMMC level requirements.