Strengthening US AI Leadership Through Collaboration with US National Laboratories

OpenAI has signed an agreement with the US's National Laboratories to begin collaborating on scientific breakthroughs using its latest reasoning models. This collaboration aligns with OpenAI's mission to ensure that AGI benefits all of humanity, and positions the US government as a key partner.

This project will be carried out in collaboration with Microsoft. Researchers will be able to use o1 or other o-series models as a shared resource on Venado, the NVIDIA supercomputer at Los Alamos National Laboratory (LANL). This infrastructure will be made available to researchers at Los Alamos, Lawrence Livermore, and Sandia National Laboratories.

Key areas of application include:

• Accelerating fundamental scientific research and deepening understanding of the principles of the universe and nature
• Identifying new approaches for treating and preventing disease
• Strengthening cybersecurity and protecting the US power grid
• Innovating energy infrastructure and maximizing the potential of natural resources
• Improving detection of natural and man-made threats, including biological and cyber threats

In addition, safety collaboration will proceed in parallel for nuclear security and CBRN (chemical, biological, radiological, and nuclear) risk management. OpenAI will provide advice on AI safety through security-cleared researchers, and plans to build more extensive safety protocols based on its existing collaborative experience in assessing risks such as bioweapon production.