The Government Accountability Office released a report in May 2026 examining how artificial intelligence is being used across federal small business contracting and innovation research programs, and where that use carries risk. The report, GAO-26-107828, found that the Small Business Administration failed to consistently meet federal transparency requirements for reporting its AI use cases, leaving a significant governance gap in one of the agencies where AI adoption in acquisition-related functions is most consequential for the small business community.

AI Use Cases in Small Business Programs

The report examined AI applications at SBA and Office of Small and Disadvantaged Business Utilization offices at other federal agencies, as well as within Small Business Innovation Research and Small Business Technology Transfer programs. At SBA and OSDBUs, GAO identified use cases including AI-assisted market research to identify broader contractor pools, tools to summarize industry trends and evaluate supplier capabilities, systems to compile acquisition forecasts and flag incomplete submissions, internal training material development, and chatbot or knowledge tools for customer support. In SBIR and STTR programs, AI is being tested for matching proposal reviewers to technical submissions, checking proposals for compliance, assisting with due diligence on awardees, detecting potential fraud indicators, and improving feedback provided to unsuccessful applicants.

Those applications address real operational challenges in programs that process high volumes of documentation, proposals, and compliance reporting. AI tools that automate routine document review, flag anomalies in submissions, match reviewers to proposals based on technical expertise, or identify statistical patterns suggesting fraud or eligibility problems have the potential to reduce processing time and improve consistency across program administration, outcomes that matter both to the agencies running the programs and to the small businesses competing for set-aside contracts and SBIR awards who depend on timely and consistent evaluation of their submissions.

Risks the Report Identified

GAO identified a range of risks associated with AI use in small business contracting contexts. Automation bias poses a risk that human reviewers will defer to AI-generated outputs without adequate scrutiny, allowing erroneous or biased recommendations to be adopted as decisions without sufficient validation. Market research AI drawing on incomplete data could systematically exclude capable firms from consideration, reducing competition in ways that harm both the government and qualified businesses that would have competed. AI systems generating reviewer assignments could introduce biased outcomes if underlying models reflect historical patterns that disadvantage certain firm types. Other identified risks include false positives and false negatives in AI outputs, hallucinated rationales in AI-generated recommendations, and cybersecurity and intellectual property concerns arising from AI tools processing sensitive proposal and contractor data.

The Transparency Gap

A central finding of the report concerns SBA failure to comply with federal AI transparency requirements. Executive Order 13960, issued in 2020, and the Advancing American AI Act, enacted in 2022, both established requirements for federal agencies to publicly inventory their AI use cases. SBA did not publish its first AI use case inventory until March 2026, more than five years after the initial reporting requirement took effect. The agency attributed the compliance failure to unclear internal roles, lack of documented procedures, and loss of institutional knowledge about reporting requirements as personnel changed over time.

The transparency gap is particularly significant given the stakes of AI use in SBA programs. When AI tools influence which contractors receive set-aside opportunities, which SBIR proposals advance to award, or which firms are flagged for fraud review, the absence of public reporting on those tools denies stakeholders the visibility needed to assess whether AI is being deployed appropriately or whether its outputs are introducing systemic patterns that affect program outcomes. GAO recommended that SBA establish formal policies and procedures for AI use case inventory reporting with clearly defined roles and responsibilities, directly addressing the governance failure the compliance gap revealed. The recommendation also encompasses broader AI governance disciplines including data quality controls, human validation requirements, and explainability standards for systems that process sensitive proposal and contractor data.

Agency-Wide AI Growth

The report documented rapid growth in federal AI adoption broadly. Federal agency AI use more than doubled from 2023 through 2024, with generative AI use increasing approximately nine-fold during the same period. That growth trajectory makes the governance and transparency questions the report raises more urgent, as the scale of AI deployment in procurement-adjacent functions is expanding faster than the policy and oversight frameworks designed to manage it. GAO has made comparable findings in oversight reports on AI adoption across defense acquisition and benefits administration programs, suggesting the governance gap at SBA reflects a broader federal pattern of AI adoption outpacing institutional capacity to document and audit how these tools are being used.

SBA paused all AI use cases in March 2025 pending a management review and policy update, and as of April 2026 the pause remained in effect except for seven pilot or pre-pilot projects testing the security, value, and performance of specific AI capabilities. The pause effectively froze SBA AI deployment while governance frameworks were being revised, though the compliance failure on inventory reporting predated the pause and reflected chronic institutional gaps rather than a temporary administrative decision made during the freeze period. The seven active pilots represent a cautious return to AI deployment under tighter management review, with each pilot required to demonstrate security compliance and operational value before broader implementation is considered.

Contractors

This article covers a GAO oversight report on federal AI use in small business contracting programs. No specific prime contractor is the subject of the report. Small business contractors participating in SBIR, STTR, or SBA set-aside programs may be affected by how AI tools are deployed in program administration at SBA and other agencies. Firms whose proposals are evaluated, whose eligibility is assessed, or whose set-aside opportunities are shaped by AI-assisted processes have a direct interest in whether the tools being used are documented, overseen, and auditable. The full report, GAO-26-107828, is available at gao.gov/products/gao-26-107828.

Sources

Artificial Intelligence: Uses and Risks for Small Business Contracting and Innovation Research (GAO-26-107828) — U.S. GAO
AI and Small Business Contracting: GAO Identifies Promise, Risk, and a Transparency Gap at SBA — FedContractPros