# USDA Flags Federal AI Oversight Gaps

At a federal AI summit on October 21, 2026, a USDA IT official said staffing reductions have left the department without enough capacity for meaningful human review of AI outputs. The shortfall comes after the September 22, 2026, deadline for safeguards under OMB M-25-21 for high-impact AI systems. The signals identify no open solicitation or specific award, but point to possible agency demand for AI risk assessments, audit and monitoring tools, governance support, and operational capacity. They also distinguish AI decision-quality and accountability controls from technical security measures such as zero trust.

- Federal contractors can position AI governance, risk-assessment, audit, and continuous-monitoring capabilities as potential support for agencies facing limited review capacity; no specific procurement opportunity is identified.
- Proposals should distinguish controls that secure systems from methods for assessing AI output quality, accountability, and human oversight.
- The M-25-21 safeguard deadline has passed. Contractors supporting high-impact AI systems should be prepared to demonstrate how their capabilities address the safeguards described in the memo.

**Jurisdictions:** federal
**Industries:** Information Technology
**Topics:** Artificial Intelligence, Cybersecurity, Regulatory Compliance
**Published:** October 23, 2026

### Government Entities
- U.S. Department of Agriculture (USDA)
- Office of Management and Budget (OMB)

### Key Quotes
> AI is really more like a utility than a capital investment.
> — Natalie Buda Smith, Library of Congress director of digital strategy

### Sources
- [Keeping a human in the AI loop is ‘untenable’ due to cuts, USDA IT official says | FedScoop](https://fedscoop.com/ai-human-in-the-loop-agency-workforces) - FedScoop
- [DOGE Workforce Cuts Made Human Oversight of Federal AI "Untenable," USDA Admits](https://www.techtimes.com/articles/329429/20261023/doge-workforce-cuts-made-human-oversight-federal-ai-untenable-usda-admits.htm) - Tech Times