AI and Small Business Contracting: GAO Identifies Promise, Risk, and a Transparency Gap at SBA
The U.S. Government Accountability Office’s May 2026 report, Artificial Intelligence: Uses and Risks for Small Business Contracting and Innovation Research, examines how artificial intelligence may affect the federal government’s small business contracting and innovation ecosystem. The report was issued under the direction of Courtney LaFountain, with staff contributions credited to Winnie Tsen, Richard Zarrella, Madeline Barch, Lena Burleson, Tarik Carter, Nathan Hanks, Daniel Horowitz, Jean Recklau, Barbara Roesmann, and Farrah Stone. GAO also convened an expert panel that included specialists from academia, industry, and former government service, whose views helped inform the report’s discussion of potential AI uses, risks, and implementation barriers.
GAO’s central finding is that AI could materially assist the Small Business Administration, agency Offices of Small and Disadvantaged Business Utilization, and agency SBIR/STTR programs, but only if adopted with appropriate governance, data controls, human review, and transparency. The report is particularly timely because federal agency use of AI has grown rapidly. GAO notes that agency AI use more than doubled from 2023 through 2024, while selected agency use of generative AI increased approximately nine-fold during the same period. Against that backdrop, the small business contracting system is beginning to consider how AI might reduce administrative burden, improve market intelligence, detect fraud, and process large volumes of information more efficiently.
For agency OSDBUs, GAO identifies several potential AI applications. AI tools could assist with market research by identifying broader pools of potential small business contractors, summarizing industry trends, or evaluating sources of supplier capability. AI could also help agencies compile acquisition forecasts, flag incomplete program office submissions, develop internal training materials, and support communications with small businesses through chatbots or internal knowledge tools. These uses could reduce staff burden and improve consistency, but GAO emphasizes that the benefits are not automatic. Market research based on incomplete data could exclude capable firms, particularly those not registered in SAM.gov or those whose Dynamic Small Business Search profiles are outdated. AI-generated recommendations also create risks of automation bias, systematic error, and overreliance on outputs that may not be traceable to reliable sources.
The report identifies similar opportunities and risks in SBIR/STTR programs. AI could help match proposal reviewers to technical submissions, check proposals for compliance, assist with due diligence, detect potential fraud indicators, and refine feedback to unsuccessful applicants. These applications are attractive because SBIR/STTR programs process large numbers of technical proposals and must guard against fraud, foreign risk, duplication, and eligibility problems. Yet GAO is appropriately cautious. AI tools can produce false positives, false negatives, hallucinated rationales, or biased reviewer assignments. More importantly, SBIR/STTR submissions often contain proprietary research and technical data. Any AI-enabled review process must therefore account for confidentiality, intellectual property protection, cybersecurity, and legal defensibility.
GAO also examines SBA’s own use of AI. SBA previously used AI, including a machine learning tool related to pandemic loan fraud risk, but paused all AI use cases in March 2025 pending management review and policy updates. SBA later began several pilot or pre-pilot use cases in 2025. However, GAO found that SBA had not consistently complied with federal requirements to publicly report AI use case inventories. Although reporting requirements have existed in various forms since Executive Order 13960 and the Advancing American AI Act, SBA did not publish its first AI use case inventory until March 2026. GAO attributes this gap partly to unclear roles, lack of documented procedures, and loss of institutional knowledge.
The report’s practical lesson is that AI adoption in small business contracting should begin with process discipline, not technology enthusiasm. Agencies should first identify where AI can produce measurable mission value, then ensure data quality, human validation, explainability, cybersecurity, and reporting compliance. GAO’s recommendation that SBA establish policies and procedures for AI use case inventory reporting is narrow, but the broader message is larger: AI may improve small business contracting, but only if agencies govern it with the same seriousness they apply to procurement integrity, transparency, and public accountability.
Disclaimer:
This article is for general informational purposes only and does not constitute legal, procurement, cybersecurity, or technology advice. Readers should review the full GAO report, applicable statutes, OMB guidance, SBA rules, and agency-specific requirements before relying on any conclusions.