OTA Prototype Competitions Have a Federal Forum: What the Blue Water Autonomy Rulings Mean for Defense Acquisition

BLUF — Bottom Line Up Front

Two October 2026 Court of Federal Claims opinions in *Blue Water Autonomy, Inc. v. United States* have changed the litigation landscape for OTA prototype competitions. The first, issued October 29, held that COFC has jurisdiction to hear protests of OTA prototype competitions connected to a procurement under 28 U.S.C. § 1491(b)(1). The second, issued October 30 and reissued November 3, reached the merits: the Navy's elimination of Saildrone in the Medium Unmanned Surface Vessel competition was arbitrary and capricious because it applied Criterion 2 evaluation standards inconsistently between Saildrone and an offeror it advanced — in a paper evaluation, not an at-sea demonstration. Program offices that have treated OTA's protest resistance as a structural feature of the vehicle need to revisit that assumption. The practical question is not whether to use OTA, but how to build the documentation discipline that judicial review now requires.

Two opinions, two questions

Judge Philip Hadji issued two separate decisions in the consolidated *Blue Water Autonomy* litigation. Understanding what each holds — and what each does not — matters for drawing operational conclusions.

October 29 (ECF 107): Jurisdiction. The court denied the government's motion to dismiss, holding that COFC has jurisdiction over protests of OTA prototype competitions when those competitions are connected to a procurement or proposed procurement under the Tucker Act, 28 U.S.C. § 1491(b)(1). The opinion rejects the government's position that OT agreements fall categorically outside the court's bid protest jurisdiction.

The ruling's relationship to prior COFC decisions is specific. The February 2025 *Raytheon v. United States* decision held that COFC can hear protests of OTAs "intended to provide the government with a direct benefit in the form of products or services." *Blue Water* builds on that reasoning. The June 2025 *Telesto Group v. United States* decision took a different approach — Judge Hertling coined the phrase "jurisdictional blackout" to describe the gap, then held that challenges to OTA prototype competitions are outside protest jurisdiction during the prototyping phase itself, with jurisdiction arising only after prototyping ends. Judge Hadji expressly declined to adopt the *Telesto* framework, creating an explicit intra-court split on the question. The Federal Circuit has not addressed OTA jurisdiction under § 1491(b)(1); no COFC trial decision is binding on another. This is a live jurisdictional question, not a settled rule.

October 30/November 3 (ECF 111): Merits. On the merits, the two protests resolved differently. For Saildrone, the court found the Navy engaged in disparate treatment under Criterion 2 (vessel solution design) of the MUSV Phase 1 paper evaluation and ordered reevaluation using consistent standards. For Blue Water Autonomy, the court denied relief; the Navy's handling of its proposal was reasonable.

The remedial order is tailored: it sets aside the Criterion 2 finding against Saildrone and bars the Navy from relying on its Criteria 3 and 4 findings, which the government itself withdrew during litigation. It does not cancel any production award, halt the Test Event for other offerors, or require Saildrone to be advanced — the reevaluation outcome is left to the Navy's technical judgment. Pending reevaluation, the court capped obligations for MUSV production at $1.5 billion of the $2.1 billion Congress appropriated for medium unmanned surface vessels (Pub. L. No. 119-21, § 20002(31), available through September 30, 2029). The production awards already made to Galliano Marine Services, HII Unmanned Systems, and Saronic Technologies — approximately $1.245 billion of the $2.1 billion appropriation — are unaffected. Prototype completion awards of $15 million each to Leidos, PAC MAR, and Sea Machines for successful completion of their prototype OTs are also left in place. The Navy has 60 days to complete the reevaluation. Whether the government will appeal has not been announced.

The MUSV Phase 1 evaluation problem

The *Blue Water* merits ruling turns on how the Navy conducted its Phase 1 paper evaluation — not on at-sea performance. Understanding the specific evaluation failure clarifies both what the court found and what program offices should take from it.

The MUSV competition used a two-phase structure. Phase 1 was a paper evaluation of written proposals; vendors who received acceptable ratings under Phase 1 criteria advanced to Phase 2 — the MUSV Test Event, involving on-water demonstrations. Twenty-five offerors submitted Phase 1 proposals; seven received acceptable ratings and advanced.

Under Criterion 2 (vessel solution design), offerors were required to complete a J-8 Placemat Template — a standardized form capturing the vessel's physical characteristics and payload capacities. The court examined whether the Navy applied its Criterion 2 standards consistently across the offeror pool.

Saildrone's Placemat listed a Maximum Displacement of 231 metric tons. Its payload-condition entries showed total displacements of 256 MT (with a 25 MT payload) and 303 MT (with two twenty-foot equivalent unit containers) — both exceeding the stated Max Displacement. The Navy eliminated Saildrone under Criterion 2 on that basis.

Sea Machines submitted two vessels — STEAMRUNNER and STEAMRACER. STEAMRUNNER's Placemat showed the same type of inconsistency Saildrone's did: its payload-condition displacements exceeded its stated Max Displacement. STEAMRACER's Placemat left the payload-condition displacement fields blank entirely. The Navy advanced both Sea Machines vessels.

The government did not dispute at oral argument that the two companies made similar errors in populating the J-8 Placemats. Its defense — that STEAMRUNNER was an existing vessel and its narrative numbers were consistent, and that "that's the technical expertise of the Navy finding those things" — did not appear in the actual evaluation record. The court found that rationale unavailable to it because the Navy had not documented it. The government acknowledged that STEAMRACER was new construction, undermining the "existing vessel" explanation for one vessel entirely.

The court held the Navy's stated basis for eliminating Saildrone cannot be reconciled with its acceptance of Sea Machines' Placemat. It did not independently determine whether Saildrone's vessel could actually meet the requirements — that judgment remains the Navy's. It held only that the Navy applied different standards to Saildrone than to an offeror it advanced.

Blue Water Autonomy's elimination, by contrast, was upheld. The court found the Navy's Criterion 4 finding — that Blue Water's catamaran test vessel and its proposed axe-bow monohull production vessel were sufficiently different that the test vessel did not validly demonstrate the production system — was reasonable. Blue Water's Criterion 4 challenge did not survive on the merits.

Why OTA's protest resistance mattered

Understanding what the *Blue Water* rulings change requires understanding why the gap mattered in the first place. GAO reaffirmed in *MD Helicopters* (B-417379, Apr. 4, 2019) that OTAs "are not procurement contracts" and that it generally does not review protests of OTA awards. GAO does retain jurisdiction to review whether the government improperly used OT authority — for example, whether a transaction labeled an OTA was actually a procurement contract — but that is a narrow question distinct from a standard bid protest of a down-select decision.

That left a practical gap: contractors who lost an OTA down-select had limited formal recourse. District court APA suits are an alternative, but they come with their own procedural constraints and are rarely fast.

Program offices in the defense autonomy and AI space have used OTA as the vehicle for rapid prototype competitions with non-traditional firms — AFWERX, DIU, and the Navy's unmanned programs office among them. Whether reduced litigation exposure was explicitly treated as a feature of that choice is a matter of program-specific judgment. The operational reality is that the protest landscape for OTA down-selects was materially narrower than for FAR-based awards, and that gap shaped how programs were structured.

The *Blue Water* rulings don't eliminate that gap entirely; GAO still doesn't hear standard OTA down-select protests. But COFC is now a credible forum for a contractor with a viable disparate treatment or misapplied criteria claim. That is a meaningful change for programs running multi-vendor prototype competitions at scale.

What program offices should change now

The *Blue Water* rulings do not make OTA the wrong vehicle for autonomous systems acquisition. The flexibility to reach non-traditional vendors, test real hardware before a production commitment, and iterate on requirements based on prototype results remains genuinely valuable. For programs acquiring unmanned platforms, autonomy software, or AI-enabled capabilities, those features often outweigh what a FAR-based acquisition offers.

What changes is the risk model. Program offices running OTA competitions should now build their evaluation processes around four specific assumptions:

Down-select criteria must be explicit, stable, and consistently applied. The *Blue Water* case was not about ambiguous criteria — it was about applying defined criteria inconsistently across the offeror pool in a paper evaluation. Criteria that are specific, documented before evaluation begins, and applied with traceable discipline reduce that exposure. The *Blue Water* record shows that undocumented technical judgment, even if exercised in good faith, cannot substitute for a written evaluation rationale that appears in the contemporaneous record.

Evaluation documentation standards should assume judicial review. Not because a protest is likely, but because the standard that survives a COFC challenge is roughly the standard that survives a GAO protest of a FAR award — a rational basis for the selection decision, evidence that criteria were applied consistently, and a selection rationale that can be read without ambiguity. Building that record from the start of evaluation costs less than reconstructing it under protest conditions.

Timeline planning should include protest contingencies. A protest that reaches COFC can seek a preliminary injunction. Not every protest succeeds, and not every injunction is granted. But the schedule assumption that OTA competitions are effectively protest-proof no longer holds, and program milestones built around that assumption are exposed. The MUSV Test Event deadline of September 30, 2026 passed while the case was in litigation — a concrete example of the scheduling exposure.

Legal counsel should review evaluation plans before they are finalized. The same legal discipline applied to a source selection plan — criteria review, scoring structure, down-select rationale — should apply to an OTA competition with a large vendor pool and a significant follow-on procurement. The underlying standards are similar; the documentation practice should be too.

What would change this analysis

The *Blue Water* rulings are single COFC trial decisions, not binding precedent across the court. The Federal Circuit has not addressed OTA jurisdiction under § 1491(b)(1). Until it does — or until Congress acts — program offices are operating under a split COFC precedent that leaves jurisdiction uncertain across different OTA structures and judges.

If the Federal Circuit explicitly holds that COFC lacks jurisdiction over OTA prototype competitions, the protest exposure would narrow sharply. If it affirms jurisdiction, the landscape becomes materially more litigious for large multi-vendor OTA competitions. If Congress amends the OTA statute to either expand or restrict judicial review, the landscape changes again. The National Defense Authorization Act has previously clarified OTA procedures; a jurisdiction fix is within scope if Congressional attention turns to it.

In the meantime, the prudent position is to build evaluation records that could survive review under either outcome, rather than betting on a future clarification that hasn't materialized.

The acquisition enterprise implications

The *Blue Water* rulings land while DoW is restructuring how it acquires autonomous systems at scale. The Defense Autonomous Warfare Group, successor to Replicator, is building an acquisition and integration structure intended to support rapid competitions, iterative prototype awards, and connections to operational deployment. DRPM-UxS (the Direct Reporting Portfolio Manager for Unmanned Systems) serves as the integrating authority across unmanned and autonomous programs. The Autonomous Warfare Command — announced September 30, 2026 and targeted for standup by October 1, 2027, pending enabling legislation — is intended to manage an unmanned systems portfolio spanning services and programs.

For that enterprise to function at the pace DoW intends, the acquisition mechanisms have to be legally durable. A protest that freezes a competition mid-evaluation disrupts program timelines regardless of its legal merit. The MUSV situation demonstrates the exposure: the obligation cap and 60-day remand impose real scheduling constraints on a program whose Test Event deadline passed while the case was pending.

Programs that build evaluation discipline into their OTA structures from the start run faster when protests don't emerge, and more defensibly when they do. The *Blue Water* rulings are a prompt to treat OTA as a mature acquisition mechanism requiring the same documentation rigor as any other — not a workaround from accountability.

Structuring OTA competitions with legally defensible evaluation records requires the same coordinated work that determines whether a program survives scrutiny at every stage: requirements analysis, acquisition strategy, solicitation development, and decision review support before the competition launches, so the evaluation rationale is built into the structure from the start. WorkForce Flow supports exactly that coordination — keeping requirements, strategy, and supporting evidence together across the acquisition lifecycle rather than reconstructing a defensible record after a challenge has arrived.