Advancing the next era of national science

American leadership has always been driven by scientific leadership. The technologies behind economic growth, national security, and better lives—from energy and computing to medicine, aerospace, and advanced manufacturing—begin with the ability to discover, build, and deploy.

Frontier AI can expand that ability. It can help compress decades of scientific progress into years, strengthen the institutions behind American innovation, and give researchers new capacity to solve problems once considered out of reach.

OpenAI is committed to bringing these capabilities to American science. We are working with the U.S. government, National Laboratories, universities, and researchers to connect frontier models with the people, supercomputers, simulations, and facilities that make American science exceptional.

Our goal is to help scientists explore more ideas, test hypotheses faster, and move from insight to validated results more quickly. That requires pairing frontier models with the tools, workflows, expertise, and infrastructure of research environments, turning AI from tool to national strategic scientific infrastructure.

As part of the U.S. Department of Energy’s Genesis Mission⁠(opens in a new window), OpenAI will support participating researchers through broad access, frontier capabilities, technical collaboration, and focused scientific campaigns. We will: 

These commitments build on our work across the national-laboratory system. In an AI Jam Session with nine National Laboratories, more than 1,000 scientists tested frontier models on domain-specific problems and provided structured feedback.

We also deployed advanced reasoning models on Venado, the supercomputer at Los Alamos National Laboratory, as a shared resource across the National Nuclear Security Administration laboratories. This work explores how frontier reasoning can support complex research alongside high-performance computing.

In bioscience, OpenAI and Los Alamos developed evaluations to study how multimodal AI systems can be used safely by scientists in realistic laboratory environments. This work can help researchers understand both the opportunities and the risks that emerge when increasingly capable systems are incorporated into consequential scientific work.

Last year, we launched our formal collaboration with Department of Energy’s Genesis Mission, which brings together DOE, its 17 National Laboratories, universities, and industry. Genesis aims to integrate AI with federal scientific data, advanced computing, experimental facilities, and expert teams—and double the productivity and impact of American research and innovation within a decade.

Our support will combine two complementary approaches: broad access, so researchers can develop practical uses for frontier models; and concentrated resources for a small number of ambitious challenges.

The initial large-scale campaigns are intended to include: 

Together, these campaigns can help establish repeatable methods to apply AI to difficult national scientific challenges.

No single company, laboratory, university, or government agency can realize this opportunity alone. 

Government can coordinate priorities and invest in shared infrastructure. Laboratories and universities can define important problems and uphold scientific standards. Industry can contribute frontier capabilities, engineering expertise, and deployment at scale. Researchers must remain at the center: defining the questions, choosing the methods, challenging the outputs, and validating the results.

The United States has repeatedly translated scientific leadership into national power. Frontier AI offers a chance to renew that model—and build an advantage grounded not in a single model or momentary lead, but in a deeper national capacity to apply frontier intelligence to important problems.

OpenAI is committed to working alongside scientists, National Laboratories, universities, and government to build that capability—and help turn frontier AI into sustained scientific and economic progress.