Security teams are spending nearly half their time mitigating AI risks.

76% of organisations report ongoing internal debate about which teams should control AI security.

51% of AI attack sources originate from North America.

21% of identified AI breaches originated from third-party applications

72% of IT leaders flagged "shadow AI" as a major risk.

21% of AI attack sources originate from South America.

32% of AI attack sources originate from Asia.

45% of identified AI breaches came from malware in models pulled from public repositories.

74% of security leaders reported an AI breach in 2024. This is up from 67% reporting the same last year.

33% of identified AI breaches originated from chatbots.

34% of AI attack sources originate from Europe.

89% of IT leaders stated that AI models in production are critical to their organisation's success.

50% of IT and cybersecurity professionals say they Deploy strict access controls for AI models to safeguard AI use for traffic monitoring/analysis are.

On average, less than a fifth (18%) of organisations' security budget is currently allocated to private 5G networks.

Nearly all IT and cybersecurity professionals said they are either currently using (62%) or planning to use (35%) AI-powered security tools for private 5G networks.

More than nine in ten organizations currently using AI security admit to facing challenges in deploying the tech to private 5G networks.

The most frequently cited challenges in deploying AI security to private 5G networks are: High costs (47%), Concerns over false positives/negatives (44%), A lack of internal expertise (37%).

Just a fifth (20%) of global organisations have a dedicated team for securing their communications networks.

44% of IT and cybersecurity professionals say they use data anonymisation techniques to safeguard AI use for traffic monitoring/analysis are.

100% of IT and cybersecurity professionals say they are either currently using (86%) or evaluating their deployment (14%) private 5G networks.