Google and OpenAI Both Commit Resources to DOE Genesis Mission National Lab Programs

On July 22, 2026, Google DeepMind and OpenAI each announced expanded resource commitments to the U.S. Department of Energy's Genesis Mission, a national program to accelerate scientific discovery using AI across all 17 DOE National Laboratories. Google committed $40 million in AI tokens and cloud credits, covering tens of thousands of lab users with Gemini for Government access [2]. OpenAI committed $4 million in Codex access to roughly 2,000 Genesis researchers plus $3 million in API support for two targeted scientific campaigns — one on high-temperature superconductors and one mapping where AI can most usefully support scientific work [3]. Both companies framed their contributions in terms of American national competitiveness and scientific leadership.

The Genesis Mission represents a deliberate effort to channel commercial AI capabilities into federally funded basic research, using national labs as a testbed. Google and OpenAI are both making significant resource bets on this model, and the outcomes — whether AI tools genuinely compress scientific timelines or primarily serve as marketing vehicles — will shape how future government-AI partnerships are structured and funded.

The DOE Genesis Mission, launched in November 2025, is a federal program designed to integrate AI into research operations across all 17 U.S. Department of Energy National Laboratories. The program invites AI companies to provide tools, compute, and expertise to researchers working on a range of scientific problems. OpenAI had established an initial partnership with national labs as early as January 2025, before Genesis Mission was formally launched, with a specific agreement with Los Alamos National Laboratory [1].

On July 22, 2026, both Google DeepMind and OpenAI published expanded Genesis Mission commitments simultaneously. Google's announcement is the larger by dollar value: $40 million in AI tokens and cloud credits distributed to Genesis Mission awardees across all 17 labs, with tens of thousands of users receiving Gemini for Government seats for one year [2]. Google also disclosed that it already operated an early access program at all 17 labs before this expanded commitment. Specific deployments cited include AlphaEvolve at Pacific Northwest National Laboratory, where it is being used to map complex mathematical systems, and Gemini at the National Laboratory of the Rockies, where it cut microscope calibration time from over 90 minutes to about 13 minutes [2].

OpenAI's Genesis Mission commitment is structured differently: $4 million in Codex access for approximately 2,000 researchers at national labs and universities, plus $3 million in API support for two large-scale scientific campaigns that can draw up to $10 million in API usage for $2.5 million spent [3]. The two campaigns target high-temperature superconductors and the construction of an "Atlas of the Machine-Accessible Frontier" — a map of where current AI can already support meaningful scientific advances. OpenAI has also deployed advanced reasoning models on the Venado supercomputer at Los Alamos as a shared resource across NNSA laboratories, and jointly developed bioscience laboratory safety evaluations with Los Alamos [3].

Both companies frame their commitments in explicitly geopolitical terms, describing AI as a tool to strengthen American scientific competitiveness. OpenAI argues that frontier AI can "compress decades of scientific progress into years" [3]; Google describes frontier AI as capable of addressing major scientific challenges while accelerating groundbreaking discoveries [2]. Neither company has provided independent verification of its headline productivity claims, and neither announcement addresses how scientific outcomes will be measured against these resource commitments.