AI-driven cyberattacks now exceed ID theft, Verizon says

Security breaches exploited via artificial intelligence became the leading cause of data violations last year, overtaking ID theft, according to Verizon's annual cybersecurity report. The US carrier stated it analyzed over 31,000 incidents and estimates that 31% of breaches originated from the exploitation of vulnerabilities in an environment increasingly shaped by the rise of AI. Verizon specifically warned of cybercriminals' ability to leverage AI to significantly accelerate the exploitation of known flaws.

The report highlights that generative AI is now being utilized at nearly every stage of an attack, from initial targeting to the development of malware and offensive tools. However, Verizon suggests that the primary impact of these technologies currently remains the automation and amplification of existing techniques rather than the emergence of entirely new attack vectors. According to the study, certain hacking groups are already exploiting over 50 AI-assisted methods, with an average of fifteen techniques used per criminal organization.

Verizon also cautioned against the emergence of new specialized AI models such as Mythos, currently being tested under Anthropic's "Project Glasswing" program. According to several experts, this model possesses advanced capabilities for identifying vulnerabilities and designing exploitation methods with unprecedented efficiency. In response to this evolution, Verizon's Chief Information Security Officer, Nasrin Rezai, believes that companies must now aggressively integrate AI into their own defense and cybersecurity frameworks.