NRC Seeks Information on AI, ML-Based Nuclear Plant Cybersecurity Solution

The Nuclear Regulatory Commission is seeking information on how artificial intelligence and machine learning can spot and stop cyberattacks at nuclear power plants.

According to a sources sought notice posted on SAM .gov, vendors must evaluate technologies and approaches for analyzing nuclear plants’ cybersecurity posture. They must then conduct research to determine if AI and ML can spot anomalies that resulted from cyberattacks.

The vendor must also use existing experimental infrastructure to conduct research activities and implement a test case, FedScoop reported Tuesday.

NRC will select a contractor that, ideally, would be able to simulate nuclear plant systems. The contractor must also show how AI and ML will integrate with operational technology, determine the effects of a cyberattack on nuclear infrastructure and develop tools to catch anomalies.

The vendors must also provide personnel who know how nuclear plants operate, have expertise in nuclear plant cybersecurity, can create simulations and are familiar with AI and ML and how they apply to IT and OT. NRC said the contractor must be able to meet requirements within 16 months of a contract award.

Interested entities have until Feb. 18 to respond.

AI can be used to spot cyber threats and malicious activity more consistently than traditional software systems. AI systems use sophisticated algorithms to detect malware, recognize malicious patterns and anticipate cyberattacks.

When paired with natural language processing and trend monitoring, AI can predict threats even before actors get the chance to adapt to new situations.