The Trump administration’s U.S. Tech Force, a programme to recruit about 1,000 engineers and technical specialists to modernize how federal agencies deploy artificial intelligence, has returned to the centre of Washington’s technology debate alongside House-passed legislation to speed up permits for AI infrastructure. The two efforts address different sides of the same problem: one tries to remove regulatory delays that slow the building of data centres and power supplies, while the other tries to close a shortage of skilled people inside government itself. Together, they signal an administration strategy that treats artificial intelligence not only as a private-sector race, but as a test of whether the federal government can build, buy and run modern technology at a pace that matches the private sector.
The Tech Force was launched in December 2025 by the Office of Personnel Management, the federal government’s human resources agency, under its director, Scott Kupor. Its stated mission is to surge elite engineering and data science talent into federal agencies to solve complex digital challenges, with a particular emphasis on accelerating the adoption of artificial intelligence across government. Participants are recruited for two-year terms, work in teams that report directly to agency leadership, and focus on high-impact projects including AI implementation, application development, data modernization and digital service delivery. The first cohort was expected to be in place by the spring of 2026, with reports at launch putting the target onboarding date at the end of March 2026. That timeline matters for understanding the programme’s current stage: it is not a proposal on paper, but a hiring drive that has already moved from announcement to recruitment.
How the US Tech Force aims to modernize federal AI deployment
The scale of the ambition is deliberately large, and the number at its heart, about 1,000 specialists in the first cohort, is small only in comparison with the size of the federal workforce. Those 1,000 people are intended to be placed where their skills are scarcest, across departments including Defense, Treasury, State, Labor and Energy, as well as the Internal Revenue Service and the Centers for Medicare and Medicaid Services. Later reporting on the recruitment pool also named the Departments of Homeland Security, Veterans Affairs and Justice among the destinations for the new technologists. The spread across civilian, financial, health and national security agencies reflects a central claim of the programme: that the government’s technology problem is not confined to one department, but is a government-wide gap in engineering, data and AI capability.
The work itself is broader than writing code. A Tech Force fellow working on AI deployment might help an agency move a system from a pilot project to a service that handles millions of transactions, clean and connect data that is currently locked in separate legacy systems, or replace a manual workflow with an automated one that can be audited and explained. In a health agency, that could mean modernizing claims or benefits systems; in a finance agency, it could mean strengthening the infrastructure that moves public money; in a defence or homeland security agency, it could mean applying AI tools to logistics, detection or analysis under strict security controls. The common thread is that the fellows are meant to work on problems that agency leaders have identified as critical, rather than on isolated experiments, and to do so with the backing of senior officials who can remove bureaucratic obstacles.
Recruitment has been pitched at a particular slice of the technology workforce. The programme targets highly skilled professionals in software engineering, AI, cybersecurity, data analytics and technical project management, with a notable emphasis on early-career technologists. One account of the launch described the target group as professionals with five to seven years of experience or fewer, a cohort that officials said was underrepresented in government, alongside engineering managers drawn from the private sector. Annual compensation was reported at launch at between 150,000 and 200,000 dollars, a range designed to be competitive with, though not necessarily equal to, private-sector pay for comparable roles. Some later accounts reported a slightly lower floor, which is a reminder that salary bands can vary by role, location and agency, and that applicants should check the specific posting rather than relying on a single headline figure.
Who is hiring, who is partnering, and how strong is the demand
The Office of Personnel Management is leading the programme, but it is not operating alone. More than twenty private-sector companies have been named as partners, including Amazon Web Services, Apple, Google, Microsoft, Nvidia, OpenAI, Oracle, Palantir and Salesforce, alongside Anduril, Dell, Meta, Adobe, ServiceNow and Elon Musk’s xAI, among others. Those companies have pledged to consider Tech Force participants for employment after their two-year government terms end, a feature that is presented as a benefit for recruits who want a route back into the private sector, and for the government, which can attract people who would not commit to a traditional forty-year federal career. Kupor has framed that point explicitly, saying the goal is not to recruit people for a lifetime in government, but to get the right talent onto the right problems for a defined period.
Demand for the limited places has been striking, at least on the administration’s own account. In December 2025, Kupor said about 25,000 people had expressed interest in joining the programme, a figure reported by Reuters with the important caveat that Reuters could not independently verify it. The interested candidates were to compete for 1,000 spots in the first cohort, a ratio that, if the interest figure is accurate, would make the programme highly selective. That caveat should be kept in mind: expressions of interest are not applications, and neither figure tells us how many candidates ultimately met the security, skills and suitability requirements for sensitive federal roles. Even so, the reported level of interest suggests that the idea of a short, high-impact government tour appeals to a segment of the technology workforce that traditional federal hiring has struggled to reach.
The partnership model also raises questions that supporters and critics answer differently. Supporters argue that bringing private-sector engineers into government for two years, with a clear return path, is the fastest way to import current skills, working methods and knowledge of modern AI tools into agencies that have historically hired slowly and retained staff for decades. Critics, by contrast, raise concerns about revolving-door risks, asking whether close ties between agencies and the companies that supply AI systems could blur the line between public service and private interest, and whether a two-year rotation can deliver lasting change once fellows leave and institutional knowledge walks out with them. Those concerns are not findings of wrongdoing against any individual or company, and they should be presented as questions in an ongoing debate, not as established facts. But they point to a real design challenge for any short-term fellowship: how to make sure the systems, documentation and skills remain in the agency after the fellows depart.
The House AI permitting bill and why it arrived alongside Tech Force
The Tech Force has come back into focus alongside the House’s passage of the SPEED Act, a bill designed to streamline federal permitting for infrastructure related to artificial intelligence projects. The legislation targets the National Environmental Policy Act, the law that requires extensive environmental reviews before major construction projects can proceed, by tightening timelines for those reviews and by reducing the period during which related lawsuits can be filed, from six years to 150 days in the accounts of the bill. Supporters, including major technology companies such as OpenAI, Microsoft and Micron, argue that those changes are needed to build data centres and their power supplies quickly enough for the United States to stay competitive with China in AI. The bill now moves to the Senate, where it faces further scrutiny and possible revision before any final vote, so its House passage was a step in the legislative process, not yet a change in law.
The pairing of the two initiatives is not accidental in policy terms. The SPEED Act addresses the physical constraint on AI growth, the time it takes to permit and build the infrastructure that AI systems run on, while the Tech Force addresses the human constraint, the shortage of engineers and data specialists inside government who can deploy AI tools responsibly and at scale. An administration that can build data centres faster but cannot staff the agencies that use AI, or that can hire engineers but cannot power the systems they build, would still fall short of its stated goal of American leadership in AI. Viewed together, the two efforts therefore represent a two-track approach: faster building on the outside, and deeper technical capacity on the inside.
That approach also sits within a wider national AI policy framework advanced by the White House, including an executive order intended to establish a national approach to AI policy and to reduce regulatory barriers to American innovation. One of the framework’s central arguments, as described in recent coverage of the lame-duck AI agenda, is that a patchwork of conflicting state laws would undermine American innovation and the country’s ability to lead in the global AI race. Whether Congress ultimately enacts a uniform national standard, and how it balances that goal against state regulation, parental controls, protections for consumers facing data-centre costs, and safeguards against AI being used to censor lawful speech, remains an open legislative question. The Tech Force does not settle those policy debates, but it does determine whether the federal government itself has the people to implement whatever rules and systems emerge from them.
What success would look like, and what remains unanswered
Judging the Tech Force will require more than counting heads. The first test is whether the 1,000 places are filled with people who have the specific skills agencies need, rather than with generalist technologists who require long training before they can contribute. The second test is whether fellows are placed on projects that matter to agency missions and to the public, and whether those projects deliver measurable improvements, such as faster processing, lower error rates, more secure systems or lower long-term costs. The third test is durability: whether agencies retain the code, data pipelines, documentation and internal expertise after a cohort leaves, so that the next cohort builds on the last rather than starting again. None of those outcomes can be assessed from announcement figures alone, and the programme’s public reporting so far has focused more on recruitment interest than on project results.
Other open questions concern scale and continuity. The 1,000-person figure has consistently been described as an initial cohort, but it is not yet clear from public reporting how large later cohorts would be, how funding would be sustained across administrations, or how the programme would be evaluated against alternatives such as direct permanent hiring, contractor support or training for existing federal staff. Security clearance, pay competitiveness for the most senior AI specialists, and the ability to work across agencies with different data rules and legacy systems are practical hurdles that every government technology programme faces, and the Tech Force will face them too. Its two-year design helps with recruitment, but it also creates a regular turnover challenge that permanent teams do not face in the same way.
For applicants and for the public, the most useful way to follow the programme is therefore to watch for concrete milestones rather than headline numbers. Those milestones include the agencies and projects to which fellows are assigned, the proportion of fellows who complete their terms, the share who move into permanent federal roles as opposed to returning to partner companies, and any published assessments of what their projects achieved. If the Tech Force can show that a short, well-supported tour in government produces systems that last, it may become a model for how the federal workforce adapts to a technology that changes faster than traditional hiring cycles. If it cannot, it will join a long list of well-intentioned fellowship schemes that attracted talented people but left the underlying systems largely unchanged. Either way, its renewed prominence alongside the House AI permitting bill shows that, in the current AI debate, the question is no longer only what technology the United States can build, but whether its government can build and use that technology itself.
Sources
USN reviewed reporting from multiple outlets for this story. Read more from Reuters, Data Center Dynamics and FedScoop.