For awareness Low Awareness
Frenos estimates two years before AI can reliably plan attacks on industrial systems
Industrial Cyber
What happened. Frenos tested six frontier AI models on over 500 simulated OT attack scenarios. Three produced workable plans in about 90% of attempts, but only about two-thirds in repeat tests. The models still needed supplied target information and skilled human oversight.
Who is affected. Critical infrastructure operators and industrial control system owners. The source does not say which organisations are affected.
What to do. Brief your team on the Frenos findings. Use the estimated two-year window to review your OT defences and asset information.
When to act. For awareness: background reading, nothing to schedule