In 1957, the Soviet Union launched Sputnik, and the United States realized that it was on the losing end of a technology race. Suddenly, the federal government had to identify, recruit, and educate the next generation of tech talent to protect American national security and solidify American competitiveness.
That panic led to a talent mobilization unlike anything seen in peacetime.
Today’s equivalent of Sputnik isn’t a satellite. It’s an AI model trained in China by an adversary that doesn’t share our freedoms, laws, or values. The United States won the space race not because we outspent the Soviets on rockets — as a percentage of GDP, the USSR spent far more. We won it because we out-recruited them on talent.
THE EXISTENTIAL THREATS TO ARTIFICIAL INTELLIGENCE
The AI race will be decided the same way — and last month, the Trump administration took two concrete steps that matter. First, an executive order that treats AI as a national security issue, including facilitating early government access to industry’s most dangerous models. Second, a National Security Presidential Memorandum that addressed the question of the moment: who in government is actually going to do this work?
The memorandum calls for a “strategic reserve of non-government AI talent,” which is a good idea with genuine bipartisan roots. As a former member of the White House AI and Tech Talent Task Force, I’ve watched this idea stall before. The Biden administration’s National Security Commission on Artificial Intelligence similarly called for a National Reserve Digital Corps, which was never launched. If set up, the government could “tap into a pool of technologists willing to contribute part of their time to public service,” and several members of that commission now serve on President Donald Trump’s AI advisory council. Those who follow AI policy debate energy, chips, and critical minerals, but talent is the most critical resource of all, and the government needs better access to it now.
After a year of cuts to government tech offices, the federal government’s reserve of AI talent is severely depleted. The U.S. Digital Service and the General Services Administration’s 18F office were dismantled, while the Department of Homeland Security’s Cybersecurity and Infrastructure Security Agency and the AI Corps, the U.S. Digital Corps, the Presidential Innovation Fellowship, and other programs have been significantly cut. Standing up a private-sector AI talent National Reserve could address the talent shortage, but without proper implementation, the idea will join the long graveyard of good intentions. Here are my five recommendations to make the talent reserve a reality.
First, compete on mission, not salary. The government will never out-pay Silicon Valley, but it doesn’t need to. The strongest recruiting pitch is the chance to do meaningful work on problems that actually matter for national security. The federal workforce demoralization of recent years has made this harder. Rebuilding that appeal starts with treating this talent as a strategic asset, not a bureaucratic headache.
Second, don’t wait for a crisis to “fix” government hiring. There are many systemic problems with government hiring that have well-documented solutions: Use direct hire authority (which lets agencies bypass several time-consuming steps in the federal hiring process) write job postings in plain language instead of federal jargon that scares off private sector candidates, and create shared certificates of eligible candidates so one agency’s vetting process unlocks hiring across all agencies — not just the one that ran the search. None of this is new. What is new is the urgency.
Third, revive the U.S. Digital Service Model. The U.S. Digital Service proved that embedding mid-career technologists who speak both tech and government inside agencies works. This isn’t an entry-level growth program. AI adoption is a cultural change that requires tech leaders with high emotional intelligence. The Reserve needs experienced people who can scope projects, push back when necessary, and operate without handholding. Organizations like the Tech Talent Project already bridge this divide, teaching technologists how to navigate the government and helping agencies recruit top-tier talent.
Fourth, meet the talent where they are. Requiring AI talent to relocate to Washington is a self-defeating constraint and a waste of valuable time. The federal government should site cleared facilities in major tech hubs such as San Francisco, Seattle, and Austin, so reservists can contribute to national security work without uprooting their lives.
BERNIE SANDERS WANTS THE GOVERNMENT TO OWN AI COMPANIES. THAT’S A BAD IDEA
Finally, make civic leave a private sector norm. Tech companies should offer employees the opportunity to serve in the Reserve and hold their jobs, like the civic leave models that some major tech companies have piloted. This isn’t just patriotism — it’s smart talent strategy. People who serve come back more motivated, more mission-oriented, and more valuable. Frame it as an expectation of responsible corporate citizenship in the AI era.
The tech capability gap in government is real, and the time to address it is now. The idea of the non-government AI talent National Reserve gives the government a vehicle to mobilize private-sector AI talent, if the administration has the will to do so.
Leah Siskind is a research fellow for artificial intelligence at the Foundation for Defense of Democracies and a former member of the U.S. Digital Service and the White House AI and Tech Talent Task Force.