The DOE Is Going All-In on AI. Traditional Scientists Are Getting the Bill.

Nuclear physicists at America's national laboratories opened their inboxes in March to find the kind of message no researcher wants to read. Their program budgets — the money that funds graduate students, postdocs, and the experiments that define careers — would be cut by roughly a third for fiscal year 2026. Not because Congress had slashed the Department of Energy's Office of Science budget. Congress actually rejected the administration's proposed 7 percent cut and held funding relatively steady. The money was being redirected internally, pulled from traditional research grants to feed a new institutional priority that DOE leadership has declared will "double the productivity and impact of American science within a decade."

That priority has a name: the Genesis Mission. And its $320 million appetite is reshaping what the Department of Energy will and won't pay for in ways that extend far beyond the nuclear physics community.

A $320 Million Bet That Traditional Grants Are Subsidizing

The Genesis Mission launched via executive order on November 24, 2025, directing DOE and its 17 national laboratories to build a shared AI-powered research platform. The vision is ambitious: integrate supercomputers, experimental facilities, AI systems, and massive scientific datasets into a unified infrastructure that can train scientific foundation models, deploy AI agents for hypothesis testing, and automate research workflows across more than 20 national challenges spanning advanced manufacturing, biotechnology, critical materials, nuclear energy, and quantum information science.

The initial tranche was $320 million, anchored by a $30 million award to Argonne National Laboratory for the Transformational AI Models Consortium and $40 million across four national labs for the American Science Cloud, which will distribute AI models and scientific data to researchers nationwide. In March 2026, DOE announced an additional $293 million request for applications under funding opportunity DE-FOA-0003612, with Phase I awards of $500,000 to $750,000 for nine-month projects and Phase II awards of $6 million to $15 million over three years. Applications are due April 28.

These are real dollars for real research, and the science behind the Genesis Mission — using AI to accelerate discovery in materials science, energy systems, and fundamental physics — is not frivolous. Under Secretary Darío Gil has emphasized that the platform aims to avoid duplication across agencies while positioning DOE's lab system as the backbone of American AI-for-science infrastructure. The challenge isn't that the investment is misguided. It's that in a flat budget environment, every dollar flowing into Genesis is a dollar that isn't flowing somewhere else.

Nuclear Physics Is Paying First

The somewhere else, right now, is nuclear physics. Within the Office of Science's nuclear physics program, operations — the fixed costs of running accelerators, detectors, and the facilities that house them — now consume more than half of the $866 million annual budget. That leaves roughly $196 million for research grants after an 18 percent cut from the previous year's levels. With Genesis expected to claim an additional 10 percent reallocation from each Office of Science program, nuclear physicists calculate that there will be zero funding available to renew the approximately one-third of ordinary grants expiring this fiscal year.

Zero is not a rounding error. It means that established research groups at universities — groups that train graduate students, employ postdoctoral researchers, and produce the publications and discoveries that justify the field's existence — face a funding cliff with no visible floor. A petition on Change.org calling for "support for research in nuclear physics" captured the community's alarm, though petitions rarely move budget lines.

The squeeze is compounded by a structural change that arrived in tandem. DOE's Office of Science announced it will merge its high energy physics and nuclear physics programs into a single combined program. The rationale offered is administrative efficiency and scientific synergy. But the nuclear physics community — which operates a smaller program than high energy physics — reads the merger differently. When a smaller discipline gets folded into a larger one, the smaller discipline's priorities tend to get absorbed rather than amplified.

Representative Deborah Ross captured the contradiction at a recent congressional hearing, noting that the administration was simultaneously proposing cuts to DOE's Office of Science budget, terminating billions in grants, and "firing world-class scientists" — while asking the same depleted workforce to execute an enormously ambitious AI initiative. The Genesis Mission doesn't exist in isolation. It exists in a budget that has been squeezed from multiple directions simultaneously.

The Broader Reallocation Pattern

Nuclear physics is the canary, not the whole mine. Every program within the Office of Science faces some version of the same pressure. The $293 million Genesis RFA covers challenges in advanced manufacturing, biotechnology, critical materials, nuclear fission and fusion, quantum information science, semiconductors, materials design, autonomous experimentation, and basic energy sciences. These are not narrow topics — they span most of what DOE's research portfolio covers. And the expectation that each program contribute a share of its budget to Genesis means the reallocation isn't surgical. It's structural.

The FY2027 budget proposal released on April 3 reinforced this trajectory. While final numbers depend on congressional action, the administration's request makes clear that AI integration is the top-line priority for DOE science funding, with traditional investigator-driven grants treated as a lower priority wherever tradeoffs must be made.

For researchers whose work doesn't naturally incorporate AI — theoretical physicists, experimentalists running long-duration measurement campaigns, scientists studying phenomena where machine learning offers marginal rather than transformative gains — the message is uncomfortable but unmistakable. The Department of Energy's definition of fundable science is shifting, and standing still is not a strategy.

What This Means for DOE Grant Seekers

The practical implications fall into three categories depending on where researchers sit relative to the AI pivot.

For researchers in AI-adjacent fields — materials scientists, computational chemists, data-intensive experimentalists — the Genesis Mission represents a genuine expansion of opportunity. The $293 million RFA is open to interdisciplinary teams from national laboratories, universities, and industry. Phase I awards are accessible even to smaller groups, and the nine-month timeline means results are expected quickly. Researchers who can credibly frame their work in terms of AI-accelerated discovery should be reading DE-FOA-0003612 carefully and preparing proposals before the April 28 deadline.

For researchers in traditional fields facing budget cuts, the short-term outlook is difficult. Nuclear physics grants are the most visible casualty, but any program where operations consume a growing share of the budget will face similar arithmetic. Researchers in these fields should be pursuing parallel strategies: maintaining current proposals through existing programs while identifying ways to connect their work to Genesis-funded infrastructure. A nuclear physicist studying neutron-rich isotopes may not be building foundation models, but if that research generates datasets that an AI system could use to improve nuclear structure predictions, there's a fundable angle that didn't exist two years ago.

For everyone in DOE's orbit, the merger of nuclear and particle physics programs signals that organizational boundaries within the Office of Science are not permanent. Researchers should track the implementation of the merger closely. New leadership structures, revised program solicitations, and reallocated review panels will create both risks and openings. Scientists who engage early in the new combined program's priority-setting processes will have more influence over where the surviving grant dollars go.

Beyond DOE, this pattern of AI-driven reallocation is not unique. NSF's emphasis on AI-ready infrastructure, NIH's growing investment in computational biology, and the Department of Defense's expanding AI autonomy budget all reflect the same federal-wide prioritization. Researchers who treat the Genesis Mission as an isolated DOE phenomenon rather than a signal of where all federal research funding is heading may find themselves unprepared when their own agency makes the same pivot.

The Risk Nobody Is Modeling

The deepest concern among scientists watching the Genesis rollout isn't about any single budget line. It's about what happens to the fields that produce the training data AI needs but don't get classified as AI research. Foundation models for materials science are only as good as the experimental measurements they're trained on. AI systems for nuclear energy optimization require decades of nuclear physics data generated by the very research groups now losing funding. Autonomous experimentation depends on domain expertise that takes years to develop in fields the current budget treats as lower priority.

If the Genesis Mission succeeds in building a powerful AI-for-science platform but the pipeline of traditional research that feeds it withers, the platform will eventually run on stale data and increasingly shallow domain knowledge. The Department of Energy hasn't published an analysis of how its traditional research investments connect to Genesis's data needs. Until it does, there's a real risk that the AI push cannibalizes the science it depends on.

For researchers navigating this rapidly shifting landscape — whether you're writing a Genesis proposal, fighting for a nuclear physics renewal, or trying to figure out where your discipline fits in the new DOE — tools like Granted can help you identify which funding opportunities align with your work and build competitive proposals that speak the language federal funders are now prioritizing.