DoD's New MOSA Reforms Aim to End Fake Modularity

By Andrew Park | 2025-11-17

This article is a follow up to "The New Acquisition Landscape: Implications for the Defense Industrial Base" [1].

For nearly two decades, DoD has pushed the Modular Open Systems Approach (MOSA), which is a strategy for designing weapon systems with interchangeable parts and open interfaces. Every major prime contractor framed itself as MOSA compliant. Lockheed Martin highlighted modularity across its naval, air, and missile portfolios. Northrop Grumman, Raytheon, Boeing, General Dynamics, L3Harris, and BAE Systems made similar claims. The messaging suggested MOSA had arrived.

Reality told a different story. Interfaces remained incomplete and proprietary. Systems branded as open were open in name only. DoD never achieved real modularity. It got static, proprietary architectures marketed as open while system developers retained full control.

The reforms announced on November 7, 2025 represent DoD's attempt to end this pattern and enforce the kind of openness that can't be sidestepped. These reforms are part of a rare convergence of bi-partisan support from both the Senate and House of Representatives to fundamentally transform defense acquisition [2][3][4]. The November 2025 Acquisition Transformation Strategy explicitly mandates that "The Department will fully implement MOSA in all acquisition strategies to the maximum extent practicable by leveraging open standards, pursuing software application programming interfaces, and asserting IP rights" [5].

This article explains what's changing, why DoD believes it's necessary now, and what it means for leaders across the Defense Industrial Base.

The Problem: What Fake Modularity Actually Looked Like

The pattern was consistent across programs. Interfaces (the technical specifications that allow different system components to work together) were described in documents but delivered in ways outside vendors couldn't actually use. Systems advertised as open required proprietary middleware or integration layers only the prime contractor controlled. Interface control documents (ICDs) arrived as static PDFs that weren't machine readable and omitted details needed for true plug and play integration. When subsystem competitions appeared open, vendors discovered they needed access to tools, test harnesses, or compilers only the prime could provide.

The results were predictable: competition remained closed, upgrades took years, and DoD stayed locked into the same contractors for modernization and sustainment.

Secretary Hegseth captured the frustration: "Imagine being able to swap parts or software of a critical munition without needing to completely redesign the missile. It's common sense, but we're not doing it" [6].

A 2020 acquisition expert observed the depth of this long-standing problem: "20+ years later, we are still having the same discussions on moving the modular open systems architectures (MOSA)...the FY 2017 requirement for MOSA is still hopes and dreams" [7].

The Solution: Machine Readable, Government Owned Interfaces

The reforms center on a straightforward requirement: machine readable and government owned interface definitions that any qualified vendor can use.

Machine readable means the interface specifications are in formats that software tools can directly parse, validate, and use to generate code. A PDF requires a human to read, interpret, and manually translate the specifications into working code. PDFs are inherently incomplete as interface specifications because they lack the precision and formal structure needed for actual integration, creating ambiguity that makes true plug and play compatibility impossible. Machine readable formats (like JSON, XML, or OpenAPI specifications) enable automated tooling to work with interface specifications directly. But machine readable alone isn't enough. The approach only works with tight schemas. A schema is like a strict form with exact rules: which fields must be filled out, what type of information goes in each field, and what answers are acceptable. Think of the difference between a vague instruction like "where do you live" versus a form with specific required fields for street address, city, state, and zip code, with the state field limited to valid two letter codes. Only when interfaces include these well defined schemas can automated tooling generate interface code, validate compatibility, and catch integration errors immediately. Without tight schemas, machine readable formats offer little improvement over PDFs. With them, software tools can automatically verify that a new component will work correctly before it's even built, eliminating the ambiguity that previously allowed prime contractors to claim vendor submissions "didn't meet spec."

The intent is to transform integration from a months long interpretation exercise controlled by the prime contractor into an automated process any qualified vendor can execute without coordination with the prime contractor.

The November 7th directive requires programs to "maximize use of Modular Open System Architectures (MOSA) for development programs moving forward by obtaining delivery of critical system interfaces with government purpose rights enable modular competition and supply chain resiliency" [8]. Government purpose rights are legal rights that allow the government to use the interface specifications and share them with any contractor, preventing the original developer from blocking competition.

The Acquisition Transformation Strategy reinforces this, mandating that programs "establish and maintain accessible repositories of interface specifications and supporting documentation in machine readable format that enable third party integration without original equipment manufacturer coordination" [8]. In other words, any qualified company should be able to access these specifications and build compatible components without needing permission or help from the original system developer.

The stated rationale is that when properly implemented, MOSA "enables cost effective and responsive modernization and sustainment of weapons systems, and allows for more competition throughout the programmatic lifecycle, avoiding vendor lock" [5]. Vendor lock occurs when DoD becomes dependent on a single contractor for upgrades and maintenance because no other company has access to the technical information needed to work on the system.

Why Now: Adaptation Speed Viewed as the Decisive Factor

Secretary Hegseth stated that "speed to capability delivery is now our organizing principle: the decisive factor in maintaining deterrence and warfighting advantage" [8].

DoD leadership believes that in modern conflict, adaptation speed determines battlefield relevance. Systems that can't be refreshed quickly become irrelevant. That assessment drives the shift toward enforceable modularity, based on the view that closed architectures break the ability to adapt rapidly.

The Ukraine conflict illustrates why DoD views adaptation speed as decisive. Ukrainian forces intercepted Russian Kinzhal hypersonic missiles using a 1970s era Patriot system upgraded through software, not hardware [9][10]. The shootdown demonstrated how software upgrades enable existing systems to counter advanced threats before new hardware can be built [11].

When Russia jammed HIMARS (High Mobility Artillery Rocket System) GPS targeting, U.S. and Ukrainian officials made software updates to both the targeting system and the rockets within the same week, described as "constant tweaking to get them to stay effective" [12][13]. U.S. Air Force programmers from the 68th Electronic Warfare Squadron reprogrammed the electronic warfare systems on F-16s donated to Ukraine in just two weeks to counter evolving Russian electromagnetic threats. The software updates created mission data files that enabled the jets' defensive systems to detect and counter Russian jamming, spoofing, and radar threats, improving the aircraft's ability to survive in contested airspace [14].

These stories illustrate why DoD views rapid adaptation as an urgent warfighting requirement. DoD believes real modularity is the technical foundation necessary to achieve that speed at scale. The assessment is that proprietary interfaces prevent rapid updates across programs, while government owned interfaces with machine readable specifications would allow multiple vendors to independently deploy improvements without waiting for the original system developer. The theory is that this approach will make the week long adaptation cycles demonstrated in Ukraine standard practice rather than exceptional achievements.

Five New Enforcement Mechanisms

The reforms establish enforcement mechanisms intended to work, which didn't exist before.

Mandatory compliance. The Acquisition Transformation Strategy states that "fully implementing Modular Open Systems Approach (MOSA) into acquisition strategies for all programs is paramount for promoting interoperability" [5]. MOSA shifts from encouraged practice to requirement.

Technical standards. Programs must "establish and maintain accessible repositories of interface specifications and supporting documentation in machine readable format that enable third party integration without original equipment manufacturer coordination, unless waived by the SAE (Service Acquisition Executive), the most senior acquisition official in each military branch, typically a three star or four star general or admiral" [8]. This aims to end the practice of delivering unusable PDF documentation.

Government data rights. The Department will implement MOSA "by leveraging open standards, pursuing software application programming interfaces (APIs: the technical specifications that allow software components to communicate), and asserting IP rights" [5]. The government will legally claim ownership of interface specifications so contractors can't block competitors by claiming proprietary information.

Government ownership of interfaces. The directive requires "obtaining delivery of critical system interfaces with government purpose rights enable modular competition and supply chain resiliency" [8]. The government now claims ownership of the technical specifications that allow different system components to connect and work together.

High level waiver authority. Any program wanting an exception to MOSA requirements must now get approval from the Service Acquisition Executive. Previously, mid level managers could grant these exceptions, making them easy to obtain. Now, every waiver requires justification to the person running acquisition for an entire service. This makes exceptions rare rather than routine.

What Changes on the Ground

DoD's policy attempts to close the loopholes that enabled fake modularity. Programs must expose machine readable and fully defined interface specifications in government repositories. Vendors wouldn't need the system developer's permission to build a compatible module. Prime contractors would lose the ability to hide integration layers behind proprietary tooling or undocumented behaviors. Program offices would gain the data rights and technical visibility needed to create genuine competition. Any exceptions to MOSA requirements must be approved by the Service Acquisition Executive, making waivers rare and visible at the highest levels.

This removes the ability to control a system through the interface layer, where fake modularity lived for years.

The success criteria are explicit. As Secretary Hegseth stated, "Success will be measured by the ability of qualified vendors to independently develop, test and integrate replacement modules at the component level throughout the system life cycle. There's no more complacency and no more monopolies" [6].

The Access Question Remains Open

The Navy has established a Collaborative Digital Environment that includes "a repository for modular system interfaces" accessible to anyone with a common access card. A Common Access Card (CAC) is the standard DoD smart card used for physical and network access, essentially a government ID card with embedded security credentials. Companies without existing DoD contracts typically cannot obtain them. How DoD will provide broader access to these interface repositories, particularly for startups, commercial tech companies, and small businesses new to defense, remains an implementation question that will need resolution as the reforms mature.

What This Means for Different Leaders

For service chiefs and senior DoD leaders: These reforms transform MOSA from aspiration to enforceable requirement. You now have mechanisms intended to break vendor lock and accelerate adaptation cycles. The policy gives you visibility and control over interface specifications that were previously hidden in contractor black boxes. SAE level waiver authority is designed to ensure programs can't quietly revert to closed architectures without your knowledge. Implementation success depends on maintaining repositories with genuine accessibility, enforcing machine readable interface standards with tight schemas, and keeping waiver authority at appropriate levels. Programs must deliver interfaces with precise, well defined schemas that enable automated validation, not just machine readable formats. The Ukraine examples demonstrate what's possible when systems can adapt in days rather than years. These reforms attempt to make that capability systematic rather than exceptional across your portfolios.

For program managers and acquisition professionals: Your programs now operate under mandatory MOSA compliance with high level waiver authority. The stated success criteria means qualified vendors can independently develop, test, and integrate replacement modules throughout the system lifecycle. Interface specifications must be machine readable, government owned, and maintained in accessible repositories. You now have the data rights and technical visibility intended to create genuine competition. Primes can no longer gate integration behind proprietary tooling or vague specifications. The policy gives you leverage to demand real openness and the authority to enforce it. But machine readable formats alone won't deliver modularity. You need to ensure interfaces include tight schemas with precise specifications that enable automated validation. Without well defined schemas, you'll get machine readable documentation that's as useless as the PDFs it replaces. This shifts your role from managing contractor dependencies to enabling a competitive ecosystem through technical transparency.

For defense tech startups and mid sized companies: These reforms open significant opportunities. The policy is designed to lower barriers to entry that previously kept you locked out. You can now compete on module quality and adaptation speed rather than incumbent relationships. The technical barrier is being removed. Success requires the ability to ship improvements continuously, integrate with government owned interfaces without prime contractor coordination, and respond rapidly to operator feedback. If you can build better modules faster than the primes, you now have a path to compete directly for components of major weapon systems. Whether the administrative barriers (repository access, credentialing) follow will determine how quickly you can capitalize on this shift. If access remains limited, lobby the services to provide broader access to these repositories. The policy creates the opening. Your engineering execution will determine whether you can seize it.

For prime contractor leaders: Prime contractors should understand that the new acquisition landscape rewards companies that help the DoD move faster, modernize more smoothly, and reduce long term sustainment burdens. Leaders who move early can put themselves in the best position to preserve their share in a market that’s about to become far more competitive.

The most urgent step is strengthening your engineering discipline. Hidden technical debt has been bleeding into DoD programs for years, and programs no longer have the patience or budget for it. Raising code quality, improving maintainability, and tightening your development practices will help you meet the performance and sustainment expectations that are now central to modern programs.

Primes need to modernize their interfaces before they’re asked to. The reforms make it clear that future systems will rely on clean, machine readable interfaces with tight schemas. Moving early shows alignment with DoD priorities, cuts integration friction, and builds trust with program offices that need to move quickly. It also positions you for a world where interfaces are no longer a barrier that protects incumbents.

DoD has made it clear that it wants more competitive primes than the handful that exist today, and MOSA is how that shift will happen. As interfaces open and entry barriers fall, more companies will compete for individual modules, and some will eventually grow into new primes. New entrants will move fast, but it will take them time to match your depth in mission understanding, certification, safety, and large scale integration. Those strengths still matter, but they must now be paired with modern engineering practices to stay competitive.

Government owned interfaces change where the advantage lies. Control of integration is no longer the lever. The edge now comes from engineering quality, rapid adaptation, and the ability to meet multi domain integration challenges like JDAC2, where sensors and weapons must connect across every service and domain. Prime contractors will need to develop these strengths intentionally if they want to stay competitive in the new environment.

You’ll also need far stronger product management capability. The new expectations place early risk on industry. You’ll be expected to use independent R&D, uncover real user needs, and produce prototypes before requirements mature. This is how companies like Anduril won major programs. It’s also the gap I explained in “Product Management Talent: The Critical Gap Agile and DevSecOps Can’t Fix” [15]. To compete, you’ll need teams capable of real product discovery so you’re building what the DoD actually values before it gets written into formal requirements. You’ll need product leadership that can scale this across your organization and make it an everyday operating rhythm rather than an isolated practice.

Primes that compete on merit will stay central if they deliver outstanding software, keep up with modernization demands, and respond quickly to evolving needs across domains. Those that raise the craftsmanship of their developers and pair it with strong interfaces will be very hard to displace. Those that don’t will see their position decline as MOSA opens the field to fast moving defense tech competitors.

These reforms raise the bar. If you invest now in engineering quality, product capability, and adaptation speed, you can preserve much of your position. If you wait, the next generation of competitors will move into that space before you realize it.

Why DoD Believes This Matters

DoD leadership views adaptation speed as the fundamental warfighting requirement of modern conflict. The Ukraine examples demonstrate the standard: software updates to Patriots that defeat hypersonic missiles, HIMARS countermeasures deployed within a week, F-16 electronic warfare systems reprogrammed in two weeks. These weren't exceptional achievements enabled by heroic workarounds. DoD believes they represent the adaptation tempo required to survive and win. Closed architectures and fake modularity prevented that tempo from becoming standard practice. Enforceable openness through machine readable, government owned interfaces with tight schemas is the technical foundation DoD believes will make rapid adaptation systematic across the force.

References

[1] Park, Andrew. "The New Acquisition Landscape: Implications for the Defense Industrial Base." LinkedIn. November 2025. https://www.linkedin.com/pulse/new-acquisition-landscape-implications-defense-industrial-andrew-park-plnoe

[2] Senator Roger Wicker (R MS), Chairman, Senate Armed Services Committee. "FORGED Act: Fostering Reform and Government Efficiency in Defense." December 2024.

[3] President Donald J. Trump. "Executive Order 14265: Modernizing Defense Acquisitions and Spurring Innovation in the Defense Industrial Base." April 9, 2025.

[4] Representatives Mike Rogers (R AL), Chairman, and Adam Smith (D WA), Ranking Member, House Armed Services Committee. "SPEED Act: Streamlining Procurement to Ensure an Efficient Defense." June 2025.

[5] Office of the Under Secretary of Defense for Acquisition and Sustainment. "Acquisition Transformation Strategy: Rebuilding the Arsenal of Freedom." Nov 2025.

[6] Secretary Hegseth, Pete. "Rebuilding the Arsenal of Freedom." Speech, Nov 7, 2025.

[7] "What is holding modular open systems back in defense?" Acquisition Talk. March 11, 2020.

[8] Secretary of War. "Memorandum: Transforming the Defense Acquisition System into the Warfighting Acquisition System to Accelerate Fielding of Urgently Needed Capabilities to Our Warriors" (including Appendix: Initial Directed Implementation Actions). Nov 7, 2025.

[9] "Ukraine Tested Software Turns US Patriot Missiles Into Drone and Cruise Missile Killers." UNITED24 Media. July 9, 2025.

[10] "Ukraine's Battle Tested Patriots Are Now Reportedly Smarter and Deadlier." UNITED24 Media. June 24, 2025.

[11] "Ukraine downed a hypersonic missile with a Patriot. What that says about the future of weapons." Defense One. September 20, 2023.

[12] "Russia's jamming of US provided rocket systems complicates Ukraine's war effort." CNN Politics. May 6, 2023.

[13] "US, Ukraine tweak HIMARS' software to counter Russian jamming efforts." Euromaidan Press. May 7, 2023.

[14] "Analysis: Ukrainian F-16s Get Latest US Air Force Electronic Warfare and Jamming Software." Kyiv Post. August 29, 2024.

[15] Park, Andrew. "Product Management Talent: The Critical Gap Agile and DevSecOps Can't Fix." LinkedIn. https://www.linkedin.com/pulse/product-management-talent-critical-gap-agile-devsecops-andrew-park-i7the

[16] "MOSA Redux: Modular Acquisition is New Again." Center for International Maritime Security. 2025.

[17] "Modular Open Systems Approach: DoD Research & Engineering, OUSD(R&E)." May 28, 2025.