Open-source AI could provide ‘geostrategic value,’ Trump AI plan says

The newly released White House Artificial Intelligence Action Plan includes a strong endorsement of open-source AI software, an embrace the Trump administration argues could provide “geostrategic” value. 

“We need to ensure America has leading open models founded on American values,” the plan states. “Open source and open-weight models could become global standards in some areas of business and in academic research worldwide. For that reason, they also have geostrategic value.” 

It continued: “While the decision of whether and how to release an open or closed model is fundamentally up to the developer, the Federal government should create a supportive environment for open models.”

The report, released Wednesday, recommends five steps to promote open-source AI, including supporting the technology through federal research investments and having the National Telecommunications and Information Administration — a Commerce Department agency that focuses on communications technology —  encourage small and medium-sized businesses to use open-source tools. 

The document suggests working with major technology providers and investing in the National Artificial Intelligence Research Resource, an initiative to develop federal government AI infrastructure. The government is also interested in looking for mechanisms to decrease the cost of compute resources for startups and researchers. 

Many technology groups — including the Open Source Initiative, the Information Technology Industry Council (ITI), and the Mozilla Foundation — have wanted the federal government to put more emphasis on open-source systems. Critics, meanwhile, have expressed concern about potential misuse and security risks. 

In response to the AI Action plan, Daniel Castro, a vice president at ITI, told FedScoop that the administration made the “right call” on supporting the development of the technology. The U.S., he argues, needs to lead in both open and closed-source models, adding that restricting open-source technology could hold back innovation. 

“Supporting open-source AI isn’t just good policy — it’s a smart strategy. It lowers barriers to entry for domestic developers, creating more innovation and competition. It accelerates the adoption of U.S.-developed tools and standards internationally, extending America’s technological influence,” Castro said. ”This positions the United States to help shape global norms around AI development, rather than having those norms dictated by others.” 

Notably, the plan doesn’t mention encouraging the federal government to use open-source AI.  Mark Surman, the president of Mozilla, has previously suggested that the government could use procurement guidelines as a way to promote open-source technology and privacy-protecting platforms. Open-source technology has struggled to find a “market niche,” he said in a 2023 interview with FedScoop. The federal government, he noted, could serve as a market for the tech and help nurture the industry overall. 

“Huge parts of the AI ecosystem are open source [and] mandating open source in procurement means that federal dollars are used in ways that accrue benefit back to the public and to the industry,” Surman said, adding that that “also means that the government doesn’t need to pay twice for stuff.” 

Whitney McNamara, a senior vice president at Beacon Global Strategies, told FedScoop that the potential of open-source AI hasn’t yet been tapped by the U.S. government.

“But there’s also plenty of cases where proprietary software is appropriate so this wouldn’t be the right forum to be overly prescriptive,” she said, “especially when organizations like [the Department of Defense] are already moving out in the spirit of the administration’s EO around leveraging commercial software and streamlining the acquisition process to quicken that adoption.”