Leadership Confidence Outpaces Cyber Preparedness: Willis Report

Boards assume ransomware outages last days; claims data shows a median 24-day outage and an average ransomware loss of $2.7 million.

Marina M Headshot
Adam121 Adobe Stock 315095274
adam121 AdobeStock_315095274

Corporate boards often express confidence in their cyber readiness. Yet recent high-profile cyber events show how fragile confidence can be when tested. In fact, a new Cyber in Focus 2025 report, by Willis, reveals losses are longer, broader and costlier than leaders expect.

“Boards often believe cyber risk is contained, but the data proves otherwise. Untested plans, weak vendor contracts, and unclear wordings are exactly where firms lose money, reputation, and regulatory standing. The cost of untested resilience shows up in lost revenue, shareholder disputes, and fines and it’s rising faster than boards expect. Ransomware simulations, vendor analytics, AI governance, and policy optimization can help bridge the gap between perception and reality,” says Peter Foster, chairman, global FINEX cyber and cyber risk solutions, Willis.

Key takeaways:

  • Boards assume ransomware outages last days; claims data shows a median 24-day outage and an average ransomware loss of $2.7 million. Every week offline means lost revenue.
  • Leaders often view vendor risk as secondary, yet ~50% of breaches start with suppliers (MSPs, SaaS, niche vendors). Weak liability, audit, and notification clauses drive cost; regulators increasingly expect proof of vendor oversight.
  • Most boards report having a plan, but only 68% tested it in the past year. Regulators and insurers are looking for evidence that controls work in practice, not policy statements alone.
  • Emerging frameworks, including the EU AI Act, evolving U.S. state rules, and new critical-infrastructure legislation in Hong Kong are raising expectations on governance, incident response, and disclosure.
  • Publicly-held companies account for 36% of total losses despite fewer incidents.
  • The largest single claim reached $331 million; Boards highlight AI’s upside, but claims already show deepfakes, synthetic IDs, and generative malware being used to commit fraud.
Page 1 of 91
Next Page