OpenAI and the U.S. Department of Energy (DOE) have signed a memorandum of understanding to strengthen their collaboration on AI and advanced computing, supporting key DOE initiatives including the Genesis Mission. This partnership is part of OpenAI for Science, an initiative aimed at accelerating discovery by integrating frontier AI models with real-world research environments, tools, and expert knowledge. The MOU establishes a formal framework for information sharing, coordination, and future project development. It builds on existing collaborations between OpenAI and DOE’s national laboratories, where frontier AI models have already been deployed to tackle high-impact scientific challenges. The agreement enables deeper technical exchange and sets the stage for follow-on projects as specific research goals emerge. This announcement coincides with the Genesis Mission event at the White House, where Kevin Weil, Vice President of OpenAI for Science, joined DOE leaders and industry partners. The mission brings together government, national labs, and private sector organizations to harness AI and advanced computing for scientific breakthroughs. OpenAI has been actively working with scientists across the DOE national lab system to understand how AI can support research, where it falls short, and how to safely integrate it into real scientific workflows. A key milestone was the 1,000 Scientist AI Jam Session—an unprecedented event across nine national labs that brought together over 1,000 researchers. Participants used frontier AI models to test domain-specific problems, evaluate model outputs, and provide feedback to shape future development. OpenAI is also collaborating with the National Nuclear Security Administration (NNSA) labs—Los Alamos, Lawrence Livermore, and Sandia—on critical scientific and technical research. As part of this, advanced reasoning models have been deployed on the Venado supercomputer at Los Alamos, serving as a shared resource for researchers across the NNSA network. These models are being used in high-performance computing environments to address complex scientific and national security challenges. Additionally, OpenAI and Los Alamos have developed new evaluation methods to assess how multimodal AI systems perform in laboratory settings. These evaluations go beyond text-based testing and focus on real-world impacts in high-stakes domains, incorporating expert oversight, rigorous study design, and a strong commitment to risk mitigation. OpenAI’s vision for science is grounded in two core beliefs: scientific progress accelerates when powerful tools meet skilled researchers, and AI can become a new kind of scientific instrument—expanding what’s possible, speeding up iteration, and turning insight into real-world impact. The DOE national labs are uniquely positioned to realize this vision, combining world-class infrastructure with deep expertise in areas like fusion energy, materials science, and climate modeling. By working closely with scientists, OpenAI aims to ensure AI is not only powerful but also safe, reliable, and aligned with the needs of real research. Today, OpenAI also submitted detailed recommendations to the White House Office of Science and Technology Policy outlining why 2026 should be recognized as a “Year of Science.” The filing emphasizes the importance of access to frontier AI models, computational resources, and real research environments to strengthen U.S. leadership in science and technology. Through this collaboration, OpenAI and the DOE are advancing a shared mission: to empower scientists with transformative tools that accelerate discovery and deliver tangible benefits to society.