Google Commits $40M to Genesis Mission AI Research

Google announced a $40 million commitment to support the Genesis Mission, a White House initiative aiming to double the pace of scientific discovery in the United States within a decade. The funding comes in the form of cloud computing credits and AI tokens for researchers at national laboratories under the U.S. Department of Energy (DOE). This initiative expands access to cutting-edge AI tools from Google DeepMind for thousands of scientists, enabling faster, more efficient research across fields like biology, materials, and energy.

What AI tools does Google offer?

Through this commitment, Google DeepMind is providing in-kind access to DOE’s Genesis Mission awardees. The AI for science portfolio includes five key models:

• AlphaEvolve - Gemini-based coding and discovery agent, designed to craft advanced algorithms.
• AlphaFold 3 - Model for predicting structures and interactions of proteins and other biomolecules.
• AlphaGenome - Tool for understanding how DNA variations, including non-coding genome parts, shape biology and disease.
• WeatherNext - Family of AI models for high-accuracy weather forecasting.
• AlphaEarth Foundations - Foundation AI model for mapping and understanding the planet with unprecedented detail.

Google is also providing Gemini for Government, a secure platform that includes seat access and tokens for tens of thousands of users at DOE national laboratories for one year. The platform supports work from the research bench to administration of specialized user facilities serving the broader scientific community.

How AI is already being used in national labs

The Genesis Mission implementation is already showing real impact. At Pacific Northwest National Laboratory (PNNL), senior researcher Dr. Henry Kvinge uses AlphaEvolve to map massive mathematical systems too complex for manual exploration. AI automatically uncovers hidden connections, accelerating discoveries that would typically take years.

“Modern mathematics relies on abstraction, but combinatorics offers concrete models that make complex geometry and algebra easier to understand. We found that systems like AlphaEvolve are extremely well suited for this search,” said Kvinge. “By leveraging the vast mathematical knowledge of LLMs, we can automate exploration from multiple angles. We are still experimenting, but the discoveries are already shaping our future research.”

Meanwhile, at the National Laboratory of the Rockies (NLR), senior materials data scientist Dr. Steven R. Spurgeon leads an autonomous materials discovery program. Using Gemini, his team slashed microscope calibration time from over 90 minutes to about 13 minutes, eight times faster, and reduced manual steps for focusing images from as many as 50 to just two.

“Our collaboration enables the development of autonomous experimental capabilities,” said Spurgeon. “By applying Gemini to our instruments, we drastically cut microscope calibration time and reduced the needed manual steps. This is time and attention we give back to the science itself, enabling autonomous workflows that observe, reason, and decide in real-time. It helps us explore parts of the materials design space that were previously impossible to reach manually.”

Long-term commitment to the national research ecosystem

This $40 million commitment expands on a previously announced early access program. Google DeepMind has opened its AI tools to all 17 DOE National Laboratories, and Google Public Sector has demonstrated how Gemini for Government can serve as the AI backbone for the DOE. At the 2026 Genesis Mission Summit at DOE, Google affirmed this long-term support to transform research across America, tackling energy, security, and fundamental science challenges.

What it means for developers and researchers in Indonesia

For Indonesia, initiatives like the Genesis Mission underscore the importance of government-industry collaboration in advancing AI-based research. Models such as AlphaFold are globally available and have been used by Indonesian researchers for protein structure studies. The availability of predictive tools and AI-powered discovery agents opens similar opportunities in other fields, such as environmental mapping, weather, and materials. While Google's current commitment is specific to DOE labs, the approach of providing a secure AI platform to thousands of researchers could serve as a blueprint for other nations looking to accelerate scientific innovation through technology.