48% of CISOs list limited visibility into AI usage as a top AI security challenge.

Fewer than 40% of internal audit leaders believe their internal audit function is adequately prepared to detect AI-enabled fraud.

51% of internal audit functions advise management on AI-related governance or policy updates.

45% of internal audit leaders identify deepfake audio or video impersonation as a leading AI-enabled fraud threat.

29% of internal audit leaders are concerned about forged contracts or legal documents created using AI.

83% of internal audit leaders expect their internal audit function to increase AI usage over the next year.

65% of internal audit leaders identify fabricated invoices or financial documents as a leading AI-enabled fraud threat.

41% of internal audit leaders are concerned about the use of AI to insert malicious code.

In the fastest cases, attackers moved from initial access to data exfiltration in 72 minutes, four times faster than the previous year.

85% of security leaders are concerned about AI-related infrastructure risk.

70% of security leaders say AI systems have more access than a human in the same role.

68% of incidents in the Automotive and Smart Mobility involve data and privacy breaches.

Organizations with over-privileged AI systems have a 76% incident rate, compared to a 17% incident rate for organizations that limit AI to only the privileges needed for the task.

38% of internal audit functions test or strengthen fraud prevention and detection.

92% of organizations have near-term AI initiatives in production infrastructure.

Relying on static credentials for AI systems correlates with a 20-percentage-point increase in incident rates.

North America accounted for 63.1% of all network-layer DDoS attacks globally, followed by the Middle East (16.1%) and Europe (13.7%).

Malicious web application and API transactions rose 128% year over year.

In the second half of 2025, ransomware attacks against Canada and the UK accounted for a combined 30% of attacks.

In the first six months of 2025 bad bot activity reached 89.2% of the total volume observed across all of 2024.