Most Cybersecurity Staff Don’t Know How Fast They Could Stop a Cyber-Attack on AI Systems

ISACA surveyed over 3,400 security and digital professionals and found significant gaps in enterprise AI governance and preparedness: 56% of respondents were unsure how quickly they could shut down compromised AI systems, 20% didn’t know who is accountable for AI applications, less than half have high confidence in investigating AI incidents, and many organizations lack human oversight of AI actions; experts warn organizations should establish clear roles, policies, processes, and guardrails before accelerating AI adoption.