Army AI Contract Dispute Could Force Disclosure Rules

Army faces lawsuit over alleged AI misuse in contract evaluation

A federal contractor is challenging the Army's use of artificial intelligence in awarding a $450 million contract, claiming AI-generated errors contaminated the evaluation process and weakened its competitive position. The case could establish precedents for how agencies must document and disclose AI involvement in procurement decisions.

TRAX International Corporation filed suit in the U.S. Court of Federal Claims after losing a White Sands Missile Range mission support services contract to Southwest Range Services. The company argues that what it calls "AI hallucinations" — fabricated references to TRAX's proposal that appeared in evaluation documents — resulted in an improper weakness assessment that damaged its bid.

The Government Accountability Office previously denied TRAX's protest, acknowledging the Army made an error but determining it did not prejudice the outcome. However, TRAX contends the Army never explained how the mistake occurred and refused to confirm whether the Source Selection Evaluation Board used AI during its review.

The transparency gap

According to the lawsuit, the Army stated that the Source Selection Authority did not use AI to evaluate proposals. But TRAX points out that the administrative record shows the SSA simply adopted findings from the SSEB report — and the Army would not clarify whether that board relied on AI tools.

"This court, of course, defers to agency judgments and discretion, not algorithms," TRAX argued in its filing.

No specific federal acquisition regulation currently governs AI use during bid evaluations. While procurement rules define decision-making authority, they do not address the extent to which agencies can delegate analysis to AI systems. This regulatory void creates uncertainty for both contractors and agencies.

Why it matters

The absence of AI-specific procurement rules means agencies face growing pressure to prove human oversight was meaningful — not perfunctory. If courts begin requiring detailed documentation of how AI recommendations were validated, agencies will need to overhaul their evaluation processes and record-keeping practices. The case also signals that contractors will increasingly demand upfront disclosure about AI use in solicitations, potentially forcing agencies to reveal their technological capabilities and evaluation methodologies before bids are submitted.

Legal experts see potential for new standards

Ira Rushing, an associate at Tully Rinckey, told Federal News Network that agencies must now demonstrate that humans meaningfully reviewed AI output. Without solid documentation, evaluations risk appearing arbitrary — the legal standard contractors need to meet to prevail in protests.

Alan Chvotkin of Protorae Law said contractors will likely begin pressing for disclosure about whether and how agencies plan to use AI during procurement. "Knowing they are using it for compliance checking is different than if they are using it for proposal evaluations," he explained.

Chvotkin added that regardless of the lawsuit's outcome, agencies should provide better upfront disclosure in solicitations about AI use and document the steps humans take to validate AI recommendations.

Discovery could reveal AI evaluation practices

If the case survives a motion to dismiss, the discovery process could expose details about the Army's AI implementation in procurement. Rushing noted this represents "a really cool opportunity" to force the government to turn over AI information, as courts in other contexts have ruled that AI use creates discoverable information.

The Army will likely argue that any AI-related errors were harmless and did not affect the outcome. But the SSEB report and other administrative records could reveal how extensively AI shaped the evaluation.

These details were first reported by Federal News Network.