When the world's at stake,
go beyond the headlines.

When everyone is accountable, who is actually responsible?

I can’t help but think of this question as I consider the flurry of new top-down activity on acquisition reform and modernization. In January 2026, the Defense Department and White House released four new initiatives: two memorandums on the defense innovation system and AI, an executive order on defense industry standards, and a new pilot program outlining no-fee commercial evaluation licenses. These initiatives are the latest in a series of acquisition and process reforms designed to transform how the Pentagon operates and does business with industry.

Although these initiatives are a step in the right direction, they share a structural weakness: By failing to define clear benchmarking standards, specify durable ownership, and reduce policy volatility, they risk increasing uncertainty for the very firms the Pentagon depends on. Unless implementation addresses those gaps, these reforms could constrain capital formation, concentrate portfolio risk, and weaken the industrial base rather than strengthen it. The question is not whether reform is necessary. It is whether these reforms, as written, will improve industry’s ability to deliver capability to the warfighter or make that task harder.

When all is said and done, these reforms will only work if implementation addresses uncertainty.

The Innovation Memo

Perhaps the most likely to broadly transform the way industry engages with the Pentagon is the memo “Transforming the Defense Innovation Ecosystem to Accelerate Warfighting Advantage.” This memo restructures the largest components of the innovation system, clarifies organizational authority, and realigns the system with departmental objectives, but overall falls flat in benchmarking and fails to address persistent funding challenges.

The memo places four innovation organizations responsible for creating, accelerating, and adopting commercial technology under the authority of the undersecretary of research and engineering and chief technology officer. Existing innovation convening groups will be replaced by the new Chief Technology Officer Action Group, which will also feature input from service-level innovation organizations.

The Defense Innovation Unit is to be designated a Field Activity alongside the Strategic Capabilities Office, and will work with the recently established Mission Engineering and Integration Activity group to better align innovation with mission objectives. This pending shift won’t meaningfully change contracting authorities or interactions with industry, but it will change their reporting requirements and realign them with the secretary’s office.

Service-level innovation organizations, whose software factories, labs, experimentation units, and rapid capabilities offices comprise the bulk of the innovation ecosystem, will be ordered to present a review of their respective innovation organizations. These reports will ideally realign and reorganize them in accordance with departmental objectives, nullify policy obstacles, increase industry engagements, and support the major innovation organizations.

The memo introduces Innovation Insertion Increments, new flexible contracting authorities for Portfolio Acquisition Executives, enabling them to have money set aside in their portfolio budgets each year for rapid capability insertion.

Similar to the reorganization of acquisition around the goals-oriented Portfolio Acquisition Executives, the innovation community will be asked to ruthlessly align around “differentiated technology, scalable products, or new ways of fighting,” and terminate programs that no longer serve a specific mission. The memo clarifies the purpose of each of these innovation organizations, in a bid to minimize overlapping authorities and conserve departmental resources.

Progress, but Overwhelming Tradeoffs

There’s plenty of good for industry in this memo. At nearly 300 innovation organizations and units, the Defense Department’s sprawling innovation ecosystem certainly needed a refresh and realignment with departmental priorities.

The memo also does a nice job differentiating between different types of innovation — technology innovation, product innovation, and operational capability innovation — which, for almost a decade, have generally been treated similarly. The realignment and redefinition of these organizations in the memo are built on these distinctions and could make interactions for industry intake for commercial products less confusing.

However, while a necessary update to the innovation ecosystem, this memo could stifle innovation through poor benchmarking and structurally inconsistent reorganization.

Firstly, the memo states a new policy of terminating innovation programs when the outcome no longer “justifies perseverance,” but fails to grapple with one of the most serious challenges to the innovation ecosystem: a lack of clear goals or benchmarks to evaluate innovative technology by Pentagon personnel. While the memo expresses an interest in pursuing acquisitions with “measurable outcomes,” it does not state whether innovation should be measured by transition rates, the time and cost of transition, or successful industry engagement. With unclear signaling about what constitutes a successful or unsuccessful program, startups looking to engage the Defense Department through innovation organizations may have difficulty raising capital.

Secondly, the Defense Department is leaving the review of service-level innovation organizations to their respective services. While service-level innovation organizations provide incredible benefits to the military writ large, as they often work directly with warfighters, the memo does not force the service leads to immediately confront and work to minimize the front door problem that exists for industry.

Finally, the memo expresses an interest in avoiding duplicative programs. But in the quest to centralize big-dollar innovation and avoid duplication, the Defense Department could over-index on funding a few companies or projects, rather than thinking about innovation as an attritable resource with similar ideas funded across multiple companies. Ironically, this introduces portfolio risk in concentration effects, and there may be fewer opportunities available for fewer companies moving forward.

The AI Memo

Congress and the Defense Department have been grappling with how to approach AI adoption and application for a decade. The Trump administration has taken several meaningful steps towards addressing integration challenges, with the memo “Accelerating America’s Military AI Dominance” being the latest step in the department’s efforts not to “win the future by sprinkling AI onto old tactics like digital pixie dust,” as Defense Secretary Pete Hegseth said, but by “discovering entirely new ways of fighting.”

This memo is successful in forcefully rearticulating the Department’s goals, addressing bureaucratic barriers, and signaling future demand to industry, but leaves unaddressed ownership problems and several other issues.

The memo outlines aggressive steps the department will take within the next year to accelerate AI adoption, encouraging personnel to focus on “eliminating bureaucratic barriers to deeper integration,” and rewarding employees who use it to reform old processes. The memo sets goals to enhance the speed of adoption and acquisition of the best models, eliminate bureaucratic barriers, further innovative competition and AI-native warfighting efforts, and preserve modular open architectures.

The core of the memo centers around seven pilot “Pace-Setting Projects,” which are intended to accelerate AI infrastructure, data, models, policies, and talent. These projects are broken down into three sections: AI for the warfighter, AI utilized for intelligence purposes, and enterprise-wide AI. These AI projects will be held to new standards — single accountable leaders, aggressive timelines, measurable outcomes, and rapid iteration where failure accelerates learning and improvement. The Chief Digital and AI Office is to be reorganized to achieve these objectives and support project leaders. It also directs the services to create federated data catalogues so the department can eventually begin training its models on classified, enterprise-wide data.

Clear Demand Signals, but Continued Ambiguity

This memo does send clear demand signals for industry solutions, explicitly stating that the Defense Department must lean on an “expanding corps of private sector partners” to achieve AI dominance, and leverage America’s comparative advantages in the AI race, from compute to capital markets and innovation. A historic Fiscal Year 2027 defense budget will enable the department to make these much-needed investments in AI capacity and other infrastructure.

The memo continues to express a new commercial-first preference, and in trying to avoid vendor lock-in, future opportunities for large and small contractors to tap in with their own capabilities seem likely. And if the department follows through by allowing cleared industry access to classified AI training data, the department could significantly accelerate AI model development.

Yet while the memo provides that much-needed demand signal, it diagnoses similar problems and proposes similar goals as prior AI strategies published in 2018 and 2023. At just six pages, it is far shorter than its predecessors. This brevity, likely intended to grant flexibility, will instead continue ongoing ownership issues and introduce other tradeoffs that may cause problems for industry.

As in the defense industry standards executive order, the implementation language is unclear: Who owns the AI integration problem set? While the memo seemingly takes the right steps in specifying that each Pace-Setting Project will be owned by “an exemplary program leader in partnership with a sponsoring organization,” it seems to keep most hiring, data standardization, and enterprise integration power diffused in the services and other components. Ultimately, if each service’s approach to adoption remains slightly different, it may remain difficult for industry to work across the enterprise.

While the memo aims to speed adoption, it emphasizes tearing down bureaucratic barriers that build trust in AI systems. But without rigorous testing, for example, civilians and warfighters may not trust the AI that they are being forced to use. Industry will still have to work hard to build trust in their system, and to prove that using AI will make their lives easier, rather than harder.

And over-indexing on enterprise AI may create budget tradeoffs at the expense of industry servicing infrastructure and other critical sectors.

The gradual adoption of cloud computing in the Defense Department is a loose but helpful analogy to AI adoption. Although GenAI.mil is a step towards normalizing AI’s use in the department, cultural attitudes towards AI are likely to change more slowly than the department or industry would like.

Defense Industry Standards Executive Order

Earlier in January, Trump released a widely publicized executive order, “Prioritizing the Warfighter in Defense Contracting,” taking aim at defense primes who cannot deliver capability on time or at cost.

While intended to be firm and constructive, this executive order is unclear about evaluation standards and thus creates significant regulatory instability. Absent additional procurement reform, it makes the Defense Department a more “impossible to deal with” customer than before.

At its core, this Arsenal of Democracy-esque executive order asserts that many defense contractors have underperformed on contracts while simultaneously pursuing “newer, more lucrative contracts, stock buy-backs, and excessive dividends to shareholders at the cost of production capacity, innovation, and on-time delivery.”

The order directs the Secretary of Defense to compile a list of transgressing contractors within 30 days and review their past performance (which has already begun). Companies determined to be noncompliant will be notified and given 15 days’ notice to submit a remediation plan. If the remediation plan is insufficient, the Defense Department may use authorities in the Defense Production Act and the Federal Acquisition Regulations and Defense Federal Acquisition Regulations Supplement to restrict certain financial activities like owner distributions and issuing dividends until the company is “able to produce a superior product, on time and on budget.”

All new or renewed contracts will contain language stating that punitive measures can be pursued “during a period of underperformance, non-compliance with the contractor’s contract, insufficient prioritization of the contract, insufficient investment, or insufficient production speed as determined by the Secretary.” Such measures include limits on executive base salaries. While the political overtones of the executive order target the traditional primes, parts of this executive order apply to public and private companies, software and hardware companies alike.

Strong Program Management, at the Expense of Industrial Base Stability

Read charitably, the executive order could encourage some industry to remain tightly committed to program schedules. Contractors who stay on track may receive a reputational boost. It may also encourage primes and other Tier-I contractors to invest in their Tier-II and Tier-III suppliers, who are often forced to reckon with inconsistent demand signals and limited cash flow.

But the executive order is intentionally punitive, and industry should be deeply concerned.

As in the AI memo, the Defense Department critically does not state which standards will be used to evaluate underperformance, or how likely remediation is to be successful. It also offers no guarantee that all industry will be treated equally. The hardware-centric language indicates that companies that sell software to the department but run into similar delays may not be held to the same standards, and it seems unlikely companies with close ties to the administration will face the same scrutiny.

Most critically, it offers no guarantee that the government will not change the terms of these deals in the future.

Without clearer evaluation standards, risk premiums may rise, and financing constraints will worsen (as in the innovation memo). It may prove even harder for smaller (and even large) firms to raise equity and secure loans to pursue struggling programs like munitions or shipbuilding. Traditional defense contractors already operate on thin margins and must fight to secure capital due to unstable demand signals. Shifting demand signals from the Defense Department further discourage large banks or venture capital firms from investing in the high-capital, difficult to produce, yet necessary technologies warfighters need.

This executive order is also a reactionary substitute for critical procurement reform. As a recent Cogs of War article points out, these defense contractors cannot realistically manage themselves out of delays and cost overruns. Many of the delays these defense contractors encounter are due to unaddressed brittleness in Tier-II and Tier-III suppliers, who receive little support from the federal government. Until these suppliers are given relief, their insecurity will continue to cause immense instability in the defense industrial base that cannot be brushed off through a simple remediation plan. This executive order places disproportionate pressure on the industry that is absolutely essential to the Pentagon’s plan to reindustrialize America.

Intellectual Property Pilot Program

Just last week, the Defense Department announced a first-of-its-kind pilot program — a pilot patent holiday for industry.

Like the AI memo, this program displays promise for startups. However, exclusivity uncertainty, the logic of extending these opportunities strictly to startups, and relief for existing intellectual property created in the labs may hamper its success.

This two-year pilot program grants industry access to 400 select government-owned intellectual property patents in energetics, materials, microelectronics, munitions, and critical minerals. Typically, a company seeking to commercialize government intellectual property must commit to a commercialization plan and first apply for a license, then negotiate terms and pay royalties. But in this pilot, industry can try commercialization first, commitment and fee-free, and decide later if they are interested in paying an official licensing fee. Accompanying the patent holiday will soon be the launch of a searchable database linking all non-secret patents originating from the 216 Defense Department labs, enabling the department to increase awareness of its intellectual property across industry, and hopefully gain a stronger return on investment for decades of research.

The announcement does not contain any language about exclusivity, indicating that many companies can hold evaluation licenses for one patent.

In a conversation with reporters at the event’s launch, Under Secretary for Research and Engineering Emil Michael indicated that the department would favor smaller startups, for whom regular commercial evaluation licenses would typically prove financially cumbersome.

A Lower Barrier to Entry, at a Cost

This pilot does a good job at lowering the barrier to commercialization for startups interested in entering the manufacturing space, thus broadening the industrial base. Startups may lack the legal support or time horizon necessary to negotiate licensing, the capital to pay for the patent itself and royalties, or the time or effort to develop this intellectual property themselves.

But, like the industry standards executive order, it introduces several funding problems. Firstly, if multiple industry partners can hold commercial evaluation licenses, startups (or other industry) may still struggle to finance these projects. Venture capital firms and banks are unlikely to extend financing opportunities to startups to manufacture a product, with no guarantee that they will receive the sole final license. Michael acknowledged this dissonance and said that selected startups would be connected with venture capital and Office of Strategic Capital funding. However, it’s doubtful that all companies could be given access to these opportunities, which disincentivizes investing time or money in a patent for startups.

In shutting the primes out, the department hampers its ability to build necessary reserve manufacturing capacity. Most primes are better equipped to operationalize this intellectual property than small startups, given their manufacturing institutional knowledge and economies of scale, and it’s a missed opportunity to give their manufacturing lines more stable demand signals.

Finally, this announcement doesn’t seem to offer any relief for startups that originated out of the national labs, whose founders can struggle for years to secure their own intellectual property developed in government.

A Real Change, or More of the Same?

From top Pentagon acquisition officials to the president, secretary of defense, and even Congress, there is a clear desire by all parts of the ecosystem to move quickly on acquisition reform and modernization. These newest initiatives, from memos on the defense innovation system and AI, to an executive order on defense industry standards, and a new pilot program outlining no-fee commercial evaluation licenses, are seemingly sincere attempts to continue that trend.

At their best, these new initiatives are rhetorically appealing and creative. At their worst, they are counterproductive to the goals of the administration: strengthen the industrial base, modernize Defense Department procedures and workforce, and invite collaboration with industry. The benefits of these initiatives, namely strong demand signals and signposting, are at risk of being eclipsed by poor benchmarking standards for innovation organizations and programs, inconsistent adoption of major initiatives across the services, poorly defined intellectual property rights, and more.

In the months ahead, the Defense Department must over-communicate standards and evaluation frameworks, further define ownership of key projects or initiatives, listen to industry feedback to change what doesn’t work, and continue to push hard on additional legislative reform and workforce improvements. And as a reminder, without Congressional appropriations flexibility, the department cannot fully take advantage of flexible contracting mechanisms like the new Innovation Insertion Increments.

Calls for reform are important in signaling to industry and the workforce the path ahead. But while these bursts of activity are interesting, without investments in support and communication, these initiatives are likely to be little more than window dressing.

Madeline Field is the assistant editor of Cogs of War, a vertical on defense technology and industrial issues brought to you by War on the Rocks.

**Please note, as a matter of house style, War on the Rocks will not use a different name for the U.S. Department of Defense until and unless the name is changed by statute by the U.S. Congress.

Image: Midjourney