"Given the $29.4 million price premium associated with [Southwest Range Services’] proposal, there is a reasonable possibility that the removal of this Weakness could have changed the award decision," the lawsuit states.
Trax International Corporation's lawsuit
Trax International Corporation asked the U.S. Court of Federal Claims in late July to direct the Army to reevaluate bids for mission support services at White Sands Missile Range. The company alleges the Army improperly relied on artificial intelligence in the source selection process, producing multiple "hallucinations" that created a false impression of Trax's proposal relative to the winning bidder, Southwest Range Services.
The complaint follows a Government Accountability Office (GAO) denial of Trax’s bid protest in May. The Army declined to comment on the ongoing litigation.
The $450 million White Sands award and the $29.4 million swing
Trax says Southwest Range Services won a contract worth roughly $450 million. The suit emphasizes a $29.4 million price premium tied to Southwest Range Services’ proposal and argues that a single identified Weakness in Trax’s proposal — which Trax says is unsupported and a product of AI error — could have altered the award decision if removed.
This site is the portfolio.
OSINTSights runs on Cloudflare Workers, D1, R2, and Vectorize, with an AI pipeline on Hetzner ARM. Nubivance designed, built, and operates it. We do the same for clients.
See what we buildAllegations of AI "hallucinations" and documentation gaps
The complaint states that records provided to GAO did not make clear whether the strengths assigned to the Southwest Range Services bid originated from a human member of the Source Selection Evaluation Board or from an AI tool. Trax further alleges the Army conceded one of the Weaknesses identified in its bid evaluation lacked supporting documentation and that the Weakness contained "made-up references to TRAX’s proposal, that no one on the [Source Selection Evaluation Board] checked.”
Those assertions form the core of Trax’s legal claim: that errors produced or introduced by an AI tool materially affected a high-dollar procurement decision and that the record does not explain the human or automated provenance of critical evaluative assertions.
Related protest: Salient CRGT, GAO, and Defense Department admissions
Trax is not the first contractor to allege wrongful use of AI in bid evaluations. The complaint notes that Salient CRGT previously alleged a Defense Department subagency used AI rather than human employees to evaluate bids; GAO dismissed that allegation in a decision dated Jan. 5, 2026. The Defense Department has also said it used AI to help evaluate bids in several recent solicitations, according to the complaint and related reporting.
In the Trax case, GAO already denied the bid protest in May; Trax has now moved the dispute into federal court.
What this means for contractors, the Defense Department, and procurement lawyers
- Contractors (Trax, Salient CRGT): Contractors will likely continue to scrutinize agency use of AI in evaluations and bring formal protests when they believe AI-generated errors affected outcomes, the complaint and precedent suggest.
- The Defense Department and Army: The department’s admission that it has used AI to help evaluate bids, coupled with alleged documentation gaps in this case, highlights tensions between adopting new tools and meeting established procurement transparency standards.
- Procurement lawyers and regulators (David Timm, Burr & Forman; OMB): Legal advisers expect more bid protests as “the ungoverned and undisclosed use of GenAI tools results in errors throughout the procurement,” David Timm wrote in a Washington Technology op-ed cited in the report. Timm told Nextgov/FCW in an email that few civilian agencies are reporting AI in bid evaluations as "high-impact," which affects whether minimum risk practices have been implemented or disclosed — a point he tied to difficulties contractors will face proving misuse.
The Trax complaint distills a practical legal question into a technical one: when AI tools contribute to evaluative judgments in high-dollar procurements, how must agencies document the source and reliability of those judgments? The company asks a federal court to revisit the award; GAO has already ruled against the protest. Absent clearer administrative disclosure about when and how AI is used — and who verified machine outputs — disputes like this one may increasingly migrate from GAO to the courts, where questions of record, provenance, and material price differentials like the $29.4 million claimed here will matter most.
