House Science Committee Hears DOE’s 'Genesis' Pitch as Members Press Over Cancellations and Oversight

WASHINGTON — The House Science, Space, and Technology Committee convened a hearing Dec. 10 to examine the Department of Energy’s newly announced Genesis mission and a recent departmental reorganization, as lawmakers praised the Biden administration’s stated goal of accelerating U.S. capabilities in artificial intelligence, quantum computing and fusion — and at the same time demanded documentation and answers about a wave of grant cancellations.

Undersecretary for Science Dario Gill, who Secretary Chris Wright designated to lead Genesis, described the initiative as a national "discovery" platform that will marshal DOE's 17 national laboratories, universities and industry partners to pair the nation's most powerful supercomputers, AI systems and quantum technologies with scientific instruments. "This mission will mobilize our 17 national laboratories, industry, and academia to build an integrated discovery platform," Gill said in his opening testimony. He compared Genesis to a Manhattan- or Apollo-scale project intended to "double the productivity and impact of U.S. research and development."

Why it matters: Members from both parties framed Genesis as a strategic response to global competition and as a vehicle for accelerating breakthroughs in energy, materials and national-security technologies. At the same time, several Democrats and some Republicans pressed Gill on recent DOE actions they say undercut that mission: the termination or suspension of hundreds of grants and projects valued at billions of dollars.

What Gill told the committee: Gill said DOE will announce initial investments to support the platform and data infrastructure, identifying roughly $320 million set aside for the American Science Cloud and the Transformational Model Consortia to curate scientific datasets and build shared computing and collaboration infrastructure. He highlighted the department’s supercomputers — Aurora, Frontier and El Capitan — and said new machines named Dutna, Mission and Vision will further the effort. He also noted a $625 million renewal for the National Quantum Initiative center and described fusion as a top priority.

"The data is the heart of the equation," Gill told Rep. Jay Obernolte when explaining how curated scientific datasets — like the Protein Data Bank for structural biology — can enable AI breakthroughs. Gill emphasized that Genesis would include open scientific efforts and protected, proprietary or classified components, saying, "Models are data too, so we will treat it with the same level of access control that we take with data."

Questions from lawmakers: Ranking Member Zoe Lofgren and others pressed the administration’s decision to hold the hearing without Secretary Wright; Lofgren said the committee has strong jurisdiction over DOE science accounts and said "there must be accountability" for what she described as consequential DOE actions. Rep. Suzanne Bonamici and other Democrats said DOE cancelled more than 220 projects and cut between $7.5 billion and $8 billion in grants that supported grid, efficiency and workforce programs and asked Gill to provide the documentation used to make those decisions. Gill said he would relay the committee’s request and noted he has not been involved in those program reviews.

Members also pressed on security and supply-chain risks, with Republican members repeatedly tying urgency to competition with China. Gill said the Genesis platform is being built "by design" with managed security: "We have a fantastic intelligence team ... security is built into everything that we do," he said, adding that Genesis will combine open collaboration where possible with restricted handling for sensitive models and datasets.

Guardrails and reproducibility: Several members sought clarity on how models trained by Genesis would be validated and how DOE would prevent model inversion or the inadvertent release of sensitive training data. Gill said national laboratories provide a ‘‘ground truth’’ because lab experiments can confirm or refute AI predictions. "When we push AI models to do science and engineering, we have the wherewithal and the instruments in our national laboratories to test whether that prediction ... is actually true," he said.

Metrics and accountability: Gill offered three broad metrics Congress can track: compute capacity (AI flops and the aspirational equivalence of hundreds of thousands of GPU equivalents), growth and curation quality of AI-ready datasets, and the number and capability of models that can credibly perform scientific and engineering tasks. He committed to providing more precise numeric targets on request and confirmed that Secretary Wright appointed him the accountable lead for Genesis inside DOE.

What’s unresolved: Gill did not provide committee staff with detailed documentation during the hearing about the grant cancellations, and several members said they would pursue those documents and additional briefings. Lawmakers also pressed the administration’s FY2026 budget choices — including proposals that members say would reduce support for particular clean-energy programs — and asked for headcounts and organizational charts to confirm how offices such as the Office of Clean Energy Demonstrations will be organized after DOE’s reorganization.

Next steps: The committee left the record open for 10 days for additional member comments and written questions. Gill pledged to relay members' requests for documentation and to provide follow-up figures on metrics and platform capacity.