U.S. Dept. of Energy seeks comment on using AI to improve electric grid and mitigate climate risks

The AI EO directs DOE to report on potential AI interventions for the electrical grid and electric power delivery in order to “strengthen[] our [U.S.] resilience against climate change impacts and build[] an equitable clean energy economy for the future.” In soliciting input for the report, the RFI identifies three focus areas: (1) grid security, reliability, and resilience; (2) infrastructure investment and development; and (3) climate risk mitigation. The RFI also includes broader policy questions related to AI deployment.

Security, reliability, and resilience of electric grid infrastructure and operations

The RFI asks how private actors, public-private partnerships, and government entities can develop and use AI to improve the efficiency and reliability of grid operations, focusing on areas such as (1) predictive maintenance, (2) load and supply balancing and demand management, and (3) flexible, intelligent power systems models and interconnection software. The RFI also asks how AI can be used to enhance grid management and resilience, including by identifying and mapping climate hazard impacts and enabling self-healing infrastructure, detection and diagnosis of anomalies, and situational awareness and response during and after disruptions.

Using AI in the electric grid and other clean energy infrastructure

DOE seeks input on how government entities and private actors can use AI to improve planning, siting, permitting, and investment in the electrical grid and related clean energy infrastructure. The RFI asks whether AI models could be used to improve and expedite siting and permitting reviews or to improve project monitoring. It asks what federal agencies can do to facilitate the use of generative AI, and how they can better align existing structured datasets with AI models and/or use AI to improve existing structured datasets. The RFI also asks about using AI to help improve energy equity by illuminating and addressing disparate impacts on marginalized communities.

Mitigating climate risks

The RFI asks whether and how AI can help strengthen U.S. climate resilience by facilitating prediction, preparation, and mitigation of climate-driven risk, including forecasting (and potentially mitigating) extreme climate events, predicting long-term climate impacts on resource adequacy and availability, and improving weather prediction models.

Other considerations

The RFI seeks information about the costs associated with adopting AI tools, systems, and related practices, how easily they can be implemented, and the potential public benefits. DOE also asks commenters to address risks associated with implementing AI systems, including potential data security or privacy violations, AI systems potentially perpetuating unlawful biases or discrimination or the possibility of disparate impacts on communities. The RFI also asks how the benefits of AI should be quantified and how DOE should handle liability that arises from AI use.

Next steps

The RFI gives stakeholders an opportunity to contribute to U.S. policy efforts at the intersection of energy policy, climate policy, and emerging AI technologies. DOE has expressly asked that commenters focus their responses on specific, actionable information. In addition, DOE encourages commenters to review information provided by its Office of Critical and Emerging Technologies, its AI Risk Management Playbook, and the Advanced Research Directions on AI for Science, Energy, and Security before submitting responses. Comments are due April 1, 2024.

Authored by Katy Milner, Mark Brennan, Ryan Thompson, and Ambia Harper.