A recent GAO report highlights both the promise and the pitfalls of deploying artificial intelligence across federal small business contracting and innovation research programs. According to a panel of experts convened by the agency, AI tools could streamline activities such as market research, proposal review, and fraud prevention — tasks that have long been plagued by inefficiency and, in some cases, exploitation. The Small Business Administration, which oversees critical programs like the Small Business Innovation Research initiative, could theoretically use AI to better analyze agency-submitted data and reduce bureaucratic bottlenecks that disadvantage legitimate small business participants.
However, the same report flags serious concerns that mirror the systemic vulnerabilities documented in procurement watchdog literature. Risks including inaccurate AI outputs, data privacy failures, and security exposures could introduce new vectors for waste and manipulation in a contracting ecosystem already riddled with them. More telling is the acknowledgment that agencies face significant barriers to AI adoption — namely, a shortage of technically qualified staff and a slow, cumbersome technology authorization process. These are not new problems; they are the structural conditions under which bloated, politically connected contractors have historically thrived, a dynamic explored at length in War Is Still A Racket. Without meaningful oversight reform, AI risks becoming yet another tool that benefits entrenched interests over the small businesses these programs were designed to serve.