From AI to analog, cybersecurity tabletop exercises look a little different this year

The report outlines best practices for modern tabletop exercises in the face of AI-accelerated threats, urging organizations to simulate AI-powered phishing, rapid CVE exploitation, deepfakes, prompt injection, and AI-driven data exfiltration while also preparing for attacks on their own AI systems. It recommends using AI to craft realistic scenarios, enforcing analog safeguards like out-of-band verification, planning for offline minimum viable operations, engaging external partners such as the FBI and CISA, and tailoring participation from SOC teams to the C-suite to improve decision-making speed, verification in low-trust environments, and overall resilience.