AI Integration in Military Contracts: A Double-Edged Sword
As the Department of Defense (DoD) accelerates its use of artificial intelligence (AI), Congress is simultaneously revising the defense acquisition process, potentially compromising the Pentagon’s ability to secure fair pricing and validate contractor claims. The 2026 National Defense Authorization Act (NDAA) aims to reduce data disclosures and testing requirements that determine the feasibility and suitability of defense contracts.
Proponents of these changes, including some lawmakers and industry leaders, argue that the reforms will “improve efficiency” and stimulate innovation. The defense industry, which views the existing acquisition process as overly bureaucratic, has voiced strong support. Similarly, tech companies advocate for changes that they believe will expedite the military’s ability to integrate innovative capabilities.
Despite the emphasis on efficiency and innovation, evidence suggests that the lack of cost controls and adequate testing, rather than data disclosure requirements, are the primary causes of delays and insufficient competition. Former acquisition leaders have pointed to these issues as significant obstacles in the defense contract process.
Restricting Cost and Pricing Transparency
The DoD primarily relies on contractors to provide certified cost and pricing data to assess the reasonableness of contract prices. Established by the Truth in Negotiations Act (TINA) in 1962, these requirements aim to prevent overcharging. Certified data includes detailed labor costs, invoices, and catalog prices, which are crucial for verifying prices in contracts with sole-source contractors.
Since the 1990s, Congress has gradually relaxed TINA’s requirements, raising the mandatory disclosure threshold from $100,000 to $2 million. The current NDAA proposes increasing this threshold to $10 million. Critics argue that these requirements are unrealistic, but contractors track and estimate related costs in their regular business operations.
Some contractors reportedly delay or deny the Pentagon access to cost and pricing data, contributing to acquisition delays. Studies reveal that most Pentagon contracts do not require such data, often due to small purchase quantities. For instance, TransDigm and Boeing have structured most of their contracts to fall below the current disclosure threshold.
The high disclosure threshold leads to unfair pricing, as evidenced by reviews of TransDigm and Boeing contracts, which revealed significant overcharges. Raising the threshold further could exacerbate these issues, especially as the defense industry becomes more consolidated and reliant on sole-source contracts.
The “Commercial” Exemption
Exemptions for commercial products and services further complicate cost scrutiny. Contractors are not required to provide certified data for commercial items, even when contract values exceed disclosure thresholds. Congress has broadened the commercial item definition to include technology developed for military use, creating a potential loophole for overcharging.
Recent AI-related contracts, like Palantir’s Maven Smart System and Anduril’s underwater drones, are classified as commercial, despite being tailored for military applications. This classification bypasses rigorous cost justification, as tech companies do not have to provide certified pricing data.
The Senate NDAA bill’s sections 823 and 824 propose further exemptions for “nontraditional defense contractors,” potentially eliminating the need for certified cost data and cost accounting standards. This broadening could allow companies like Microsoft and Anduril to benefit from these exemptions.
Contracting Shortcuts and Oversight Gaps
The Senate NDAA also proposes easing the use of rapid acquisition pathways, allowing the DoD to bypass competition standards for emerging technologies. The proposed changes would enable the DoD to enter into production contracts without comprehensive testing, raising concerns about committing to unproven technologies.
For instance, the Army’s ongoing struggle with the Integrated Visual Augmentation System (IVAS) project highlights the risks of insufficient testing. Despite significant investment, the project remains stalled, with unresolved user acceptance issues.
The Senate bill also suggests an “alternative test and evaluation pathway,” which could dilute rigorous testing processes by relying heavily on industry-conducted testing. This approach may blur developmental and operational testing, complicating the identification of technological failures.
These reforms, coupled with staffing cuts at the Pentagon’s testing office, risk increasing operational failures, as noted by retired military leaders.
The Path Forward
These legislative changes threaten the DoD’s ability to ensure fair pricing and competitive procurement of effective technology. Congress should reconsider these proposals and strengthen oversight of AI contracts, maintaining robust testing processes and negotiating favorable contract terms.
To prevent overcharging and ensure effective technology acquisition, Congress must address the erosion of cost and pricing disclosures, particularly in AI procurement. Reducing the disclosure threshold and narrowing the definition of commercial items are crucial steps in this direction.
While challenging, enhancing pricing and testing safeguards is essential to protect taxpayers and ensure military readiness by equipping soldiers with proven technology.
