The Trump administration's recently released "Winning the Race: AI Action Plan" places a significant, albeit contradictory, bet on open-source artificial intelligence. While the plan explicitly aims for "unchallenged global technological dominance" as a national security imperative, it simultaneously advocates for making "open-source and open-weight AI models... freely available by developers for anyone in the world to download and modify.

Article illustration 1

traffic_analyzer / Getty Images

This inherent tension – promoting unrestricted global access while seeking unchallenged U.S. supremacy – underscores a fundamental challenge. The plan rightly frames open-source AI (built on frameworks like TensorFlow and PyTorch) as a "vital catalyst for innovation," empowering startups, academics, and agencies by providing access to code and model weights. Chris Wolfe, Broadcom's global head of AI, reinforces this, stating, "Thanks to open source, the pace of AI innovation now surpasses anything closed models could have achieved." Strategically, open models are positioned to set global technical standards and counter China's influence, particularly through initiatives like DeepSeek.

However, the plan's concrete support mechanisms are notably thin. Its centerpiece for enabling open-source access hinges on the National AI Research Resource (NAIRR) pilot program, a collaboration between NIST, NSF, the Office of Science and Technology Policy, and tech companies. NAIRR is designed to provide researchers and startups with crucial computational resources, datasets, and tools. Yet, its funding is precarious. Currently operating on a $30 million federal seed grant from the Biden era and a matching $30 million in corporate donations (primarily from Nvidia), NAIRR lacks a dedicated federal funding stream. This severe limitation forced it to reject over 115 proposals, awarding resources to only 35 projects out of more than 150 submissions. Early hopes for $500 million in dedicated infrastructure funding within the recent budget bill evaporated, leaving NAIRR critically under-resourced.

Furthermore, the limited AI funding available through the "Big Beautiful Bill" comes laden with restrictions. Companies must navigate "strict domestic content rules" and prohibitions against involvement by "prohibited foreign entities." As law firm Ropes & Gray warns, this necessitates heavy investment in "due diligence, supply chain transparency, and ongoing monitoring" to avoid losing benefits or facing penalties. Licensing deals, mergers, and restructuring must be meticulously structured for compliance. These burdens disproportionately impact smaller open-source initiatives and startups seeking federal support.

The plan's call for open models "founded on American values" adds another layer of ambiguity. While specifics are absent, the document explicitly excludes support for projects involving "misinformation, Diversity, Equity, and Inclusion, and climate change," raising concerns about potential ideological gatekeeping for publicly supported AI development. Oversight of these open-source promotion efforts falls to Marco Rubio, who simultaneously holds the roles of Secretary of State, National Security Advisor, and Acting Archivist, despite lacking any documented technical expertise in AI or open-source development.

While the plan avoids mandating open-source releases, leaving decisions to developers, and promises encouragement via the Department of Commerce's NTIA for SMB adoption, it offers no actionable details. The stark contrast between the plan's bold rhetoric championing open source as a strategic weapon and its lack of substantial funding, coupled with complex compliance hurdles and vague value definitions, paints a clear picture. For the open-source AI community hoping for robust federal partnership, the AI Action Plan delivers largely aspirational support – strong on vision, but critically weak on the resources and clarity needed to truly empower innovation at the scale it envisions. The success of its open-source ambitions now rests heavily on the private sector filling the void left by inadequate and conditional public investment.