The Idaho National Laboratory (INL) has partnered with Amazon Web Services (AWS) to accelerate the development of AI-powered tools for nuclear energy projects, including the ambitious goal of creating autonomous nuclear reactors. Announced on July 23, 2025, this collaboration aims to leverage AWS’s cloud computing infrastructure and advanced AI capabilities to modernize nuclear energy research and development.
Accelerating Nuclear Innovation with AI and Cloud Computing
The collaboration between INL and AWS signifies a major step in integrating advanced AI technologies into nuclear energy research. INL, a leading national laboratory in nuclear energy research, is developing a suite of technologies that use AI to reduce the costs and timeframes associated with designing, licensing, building, and operating nuclear facilities.
The Role of Digital Twins in Autonomous Reactors
A key aspect of this partnership involves using AWS Compute and AI tools to develop a digital twin of a small modular reactor (SMR). SMRs are advanced nuclear reactors that typically range from 20 to 300 megawatts of electricity. This digital twin will serve as a virtual model of a physical reactor, utilizing near real-time data to enable advanced modeling and simulation. This capability is crucial for understanding how a physical reactor will function and is a vital step toward the autonomous operation of nuclear facilities, aiming for safer, smarter, and more responsive civilian nuclear operations.
AWS Technologies Powering Nuclear AI
AWS is providing INL with access to powerful AI and computing technologies, including:
- AI models and GPUs: Essential for processing the vast amounts of data involved in nuclear research.
- Amazon Bedrock: A service that offers secure and flexible tools for developing generative AI applications, enabling INL researchers to build nuclear energy applications using leading foundation models.
- Amazon SageMaker: A fully managed service that assists data scientists and engineers in building, customizing, and deploying foundational models.
- Customized chips: Such as Inferentia and Trainium, to support mission requirements.
This access to cutting-edge cloud computing and AI solutions will enable “nuclear energy AI at scale,” according to Chris Ritter, division director of Scientific Computing and AI at INL.
Broader Implications and Federal Initiatives
The partnership aligns with a larger INL strategy to foster an ecosystem where Department of Energy (DOE) laboratories, AI technology companies, and nuclear energy developers can collaborate. The development of nuclear energy solutions is also aimed at offering sustainable power options for data centers that handle large volumes of compute, as the demand for energy to power AI-fueled data centers continues to grow.
The announcement of this collaboration closely followed President Donald Trump’s signing of three new executive orders on AI policy, one of which focuses on scaling a strong AI infrastructure within the U.S. Although the INL-AWS partnership is not a direct result of these specific executive orders, it resonates with the broader federal strategy to accelerate AI infrastructure deployment and leverage federal lands for energy and data infrastructure development.
The U.S. Department of Energy has identified four federally owned sites, including the Idaho National Laboratory, as potential locations for private-sector development of AI data centers and associated energy projects. These sites are positioned to host new data centers and power generation, including nuclear energy, to bolster grid reliability, strengthen national security, and reduce energy costs.
Other National Laboratories Advancing Nuclear AI
While INL leads this specific initiative with AWS for autonomous reactors, other national laboratories are also actively integrating AI into nuclear energy:
- Argonne National Laboratory: Has a long legacy in nuclear innovation and is leveraging AI to improve efficiency and predictive maintenance in nuclear power plants, potentially saving the industry over $500 million annually. Argonne engineers are exploring how AI, particularly large language models, can enhance how nuclear power plant operators handle complex diagnostic information.
- Oak Ridge National Laboratory (ORNL): Is collaborating with AI company Atomic Canyon to streamline the licensing process for nuclear power plants using AI, aiming to save significant time and labor. ORNL also operates the Frontier exascale supercomputer, which is crucial for high-fidelity simulations and training AI models with vast nuclear documentation.
- Lawrence Livermore National Laboratory (LLNL): Is teaming up with AWS to develop AI tools to advance its fusion energy efforts, focusing on real-time solutions to anomalies and predictive maintenance.
- Pacific Northwest National Laboratory (PNNL): Utilizes AWS GovCloud (US) for its data processing needs, including climate modeling, and is broadly engaged in applying AI/ML, including generative AI, to accelerate scientific discovery across various domains.
These collaborations highlight a nationwide effort to integrate advanced computing and artificial intelligence to revolutionize the nuclear energy sector, making it more efficient, safer, and capable of meeting future energy demands.
