Industrial operators are adopting AI defenses faster than they are testing them, and the numbers make the gap concrete: 78% of security leaders using AI report high confidence, yet only 29% run continuous simulation testing and 44% test twice a year, less often, or not at all. Jason Allnut, OT and cyber subject matter expert at SimSpace, works through those findings on the PSDcast from Power Systems Design, where AI’s promise for industrial security sits next to its risks. Confidence built on infrequent testing is not readiness. In OT environments, where a failed response has physical consequences, that means exercising people, tools and AI together in a replica of the plant network.
