DOE Awards $68 Million for AI-Powered Scientific Research Across 43 Teams

The Department of Energy announced $68 million in funding for artificial intelligence research spread across 11 multi-institution projects and 43 individual awards, establishing one of the largest coordinated federal investments in AI for scientific discovery.

The awards, selected through competitive peer review under the DOE's "Advancements in Artificial Intelligence for Science" funding opportunity, will run for up to three years. Initial funding of $20 million comes from Fiscal Year 2024 appropriations, with remaining years contingent on congressional approval.

Foundation Models, Privacy, and Energy Efficiency

The research agenda targets three frontiers. First, teams will study how large foundation models for science improve as they scale in size and complexity — a question with direct implications for whether smaller research institutions can compete with well-resourced labs.

Second, several projects focus on privacy-preserving federated learning: training models across multiple institutions without exposing sensitive data. Argonne National Laboratory is leading work on tensor-compressed sustainable pre-training of foundation models, aiming to make AI both more trustworthy and energy-efficient.

Third, researchers will develop energy-efficient algorithms that leverage next-generation microtechnologies, addressing the growing concern that AI's computational appetite undermines its scientific utility.

Who Should Pay Attention

The funding implements directives from Executive Order 14110 on safe, secure, and trustworthy AI development. While this particular round is awarded, it signals DOE's Office of Science will continue investing in AI-for-science infrastructure — researchers at national labs and universities working on scientific computing, materials science, climate modeling, and high-energy physics should anticipate future solicitations.

The Advanced Scientific Computing Research (ASCR) program homepage lists the complete project roster and participating institutions. Researchers developing AI tools for laboratory automation, computational chemistry, or cross-institutional data sharing are well-positioned for follow-on funding cycles.

For in-depth analysis of AI funding trends, visit the Granted blog.