First-party fraud accounted for less than 10% of fraud in Latin America.

The attack rate at login jumped 216%.

18% of boards and executive teams are extremely confident in their ability to detect API attacks leveraging Generative AI

Social engineering drove 88% of material losses in the first half of 2025 in Resilience's healthcare portfolio, making human error the industry's single most consequential vulnerability.

40% of MSPs report that clients ask them to define or review BYOD policies

50% of MSPs report that clients want help implementing technical controls for BYOD

32% of digital trust professionals note that no disclosure requirements exist when AI is used to create or substantially assist with work products.

28% of incidents originate entirely outside repositories—in Slack, Jira, Confluence, and similar tools.

93% of CNI organisations experienced a cyber attack in the past 12 months

50% of CNI organisations said IT disruption/outage was the top impact of a cyber attack.

35% of CNI organisations now cite regulation as the primary driver of maturity.

32% of respondents from CNI organisatins cited the complexity of cloud environments as the root causes of cyber incidents.

By 2027, 30% of organizations will require comprehensive sovereignty of their cloud security controls to address continued geopolitical turmoil.

52% of organizations say their average ransomware payout exceeds their annual cybersecurity budget.

Over 9 in 10 (92%) IT decision makers in the U.S. are concerned about the impact of cyberwarfare on their organization as a whole.

U.S. respondents (83%) were most likely to agree with the statement: “My organization has allocated sufficient budget for cybersecurity programs, including people and processes.”

DDoS attack peak volume increased from 2.2 Tbps to 12 Tbps, a sixfold increase.

Network-layer attacks accounted for 82% of all observed DDoS incidents, a 20% increase from the previous report.

93% of consumers say it's important their provider offers cyber security.

Consumers are ten times more worried about getting bad AI advice than they are about a tool's actual cyber threat risk.