(Original Caption) Port San Luis, California: Bill Musial, of San Luis Obisbo, sits on rock overlooking the Diablo Canyon nuclear plant, September 16. Only a handful of demonstrators were on hand, September 23, and a small number of arrests were made.
San Luis Obispo, California – California’s last operating nuclear power plant, Diablo Canyon, is taking a bold step into the future by integrating artificial intelligence into its day-to-day operations. While the AI is currently used only for document retrieval, the move has sparked widespread debate among lawmakers and community watchdogs concerned about increasing automation in a high-stakes sector.
Located on California’s central coast in San Luis Obispo, the Diablo Canyon Power Plant is operated by Pacific Gas & Electric (PG&E) and remains a critical piece of the state’s energy infrastructure. Despite being scheduled for decommissioning in 2029, the plant still provides nearly 9% of California’s electricity and 17% of its zero-carbon energy.
In a partnership announced last year, PG&E teamed up with local AI startup Atomic Canyon to deploy “Neutron Enterprise,” a generative AI tool powered by eight NVIDIA H100 processors—some of the most advanced GPUs available. The AI is being used to accelerate the retrieval of technical documentation, a process that traditionally consumes thousands of staff hours annually.
“Federal and state regulations require utilities that operate nuclear power plants to manage billions of pages of technical documentation,” PG&E said in a press release. “Power plant personnel must spend both time and resources to retrieve this essential data accurately and reliably.”
Neutron Enterprise integrates with the plant’s systems using cutting-edge AI technologies, including optical character recognition (OCR) and retrieval-augmented generation (RAG), to significantly reduce document search times—from hours to mere seconds. PG&E says the AI will function as a “copilot,” not a decision-maker.
“We probably spend about 15,000 hours a year searching through our multiple databases and records and procedures,” said Maureen Zawalick, Vice President of Business and Technical Services at Diablo Canyon. “And that’s going to shrink that time way down.”
However, the initiative has drawn concern from experts and lawmakers wary of expanding AI’s role in such a sensitive industry.
“The idea that you could just use generative AI for one specific kind of task at the nuclear power plant and then call it a day—I don’t really trust that it would stop there,” said Tamara Kneese, director of the Climate, Technology, and Justice program at nonprofit Data & Society.
While current AI use at Diablo Canyon poses minimal risk, critics argue that without regulatory guardrails, its expansion could lead to overreliance on automation in nuclear settings. Under the Trump administration, federal efforts to regulate AI have slowed, with a focus on encouraging innovation rather than imposing restrictions.
Still, PG&E maintains that the technology is being deployed with caution and transparency. Full deployment of Neutron Enterprise is expected by the third quarter of 2025.
With data centers driving up energy demand and Silicon Valley leaders like Bill Gates and Sam Altman championing nuclear power as a climate solution, AI and nuclear energy are becoming increasingly intertwined. Diablo Canyon’s experiment may set a precedent not just for California, but for the future of energy nationwide.
