IT Home On October 13, it was reported that the nuclear energy industry has always been a highly cautious and risk-averse sector. Since accidents or mistakes can lead to catastrophic consequences, it is generally not appropriate to deploy new technologies with unknown risks. However, reality tells a completely different story.
According to a report by IEEE Spectrum last week, several companies have begun providing AI solutions for nuclear power.This year, almost all 94 nuclear reactors in the United States have been given the opportunity to integrate AI into their operations, and most have accepted the deployment.
In August this year, Atomic Canyon, a company based in California, USA, launched the NIVA nuclear industry virtual assistant. This system was developed in collaboration with the nuclear industry community. The virtual assistant was piloted at nuclear power plants operated by Constellation Energy, and it is now available for all reactors in the United States.
Meanwhile, the nuclear AI startup Nuclearn said that its products have been partnered with more than 65 U.S. partners, including many traditional nuclear power plants, and have also been integrated with NuScale Power Corp., which is focused on the development of advanced small nuclear reactors.
Microsoft and NVIDIA have collaborated on a project aimed at applying AI to the “entire life cycle of nuclear power plants, including site approval, design, construction, and continuous operation”.
Jerrold Vincent, Chief Financial Officer and Co-founder of Nuclearn, said there are two reasons for the rapid adoption of AI in the nuclear energy field:
First of all, those who place their future in AI are actually also keen on nuclear energy, as it provides more power for data centers.
Secondly, AI is also highly suitable for solving some of the most urgent problems in the nuclear industry. Many of the reasons that make reactors difficult to operate — complex regulations, as well as complex management and maintenance systems — are exactly the tasks that AI is good at.
According to reports, the main output of nuclear reactors (besides electricity) is administrative work. Nuclear power plants produce a large number of documents to prove compliance and record responses to any potential issues. Nuclearn’s AI product helps nuclear power plants search vast databases more efficiently, allowing them to submit documents to the U.S. government faster, or to find out how a past issue was resolved. It can be regarded as ChatGPT for nuclear energy.
According to reports from IT Home, another major factor is the obsolescence of knowledge in the nuclear energy industry. As U.S. physical industries stagnate, there are fewer and fewer scientists or technicians entering the nuclear energy field. As a result, the U.S. nuclear power industry has a skilled workforce, but the number of people is rapidly decreasing. To revitalize the nuclear energy industry, an AI capable of searching through the experience accumulated by the nuclear energy industry over decades is essential. After Atomic Canyon won the NIVA contract in 2025, the first step was to download 53 million pages of public data from the Nuclear Regulatory Commission to train the model and understand nuclear energy operations, terminology, and relevant regulations.
However, Trey Lauderdale, the CEO of Atomic Canyon, believes that, both in the short and long term, AI will not make decisions related to nuclear power plant operations, as this involves regulations and risks. “You don’t want a machine to make decisions that could affect operations. So, I believe humans will continue to be involved in nuclear power plant operations for a very, very long time.”