Human Error Drives 85% of Cyber Losses in H1 2026 as AI Amplifies Traditional Threats
Event summary
- 85.3% of cyber losses in H1 2026 stemmed from human error, up from 17.7% in H1 2024.
- No incurred losses from AI-specific attack vectors observed in the period.
- Ransomware accounted for 73% of financial losses but only 5.8% of total claims.
- Vendor-related losses dropped to 2.3%, down from 33.5% in H1 2025.
The big picture
Resilience's data reveals that while AI is reshaping cyber risk, its current financial impact comes through enhancing traditional attack methods rather than creating new ones. The sharp rise in losses from human error underscores the need for organizations to strengthen foundational cybersecurity controls as they prepare for more sophisticated AI-driven threats. The decline in vendor-related losses suggests improved third-party risk management, but recent incidents highlight ongoing operational risks.
What we're watching
- AI Threat Evolution
- How AI will increasingly amplify traditional attack methods and whether new AI-native threats emerge as significant loss drivers.
- Ransomware Defense
- The effectiveness of immutable backups in mitigating ransomware losses and the pace at which organizations adopt advanced containment strategies.
- Human Error Mitigation
- Whether organizations can implement layered controls to limit financial impact from inevitable human mistakes, especially as AI enhances deception tactics.