AI Now Powers 84% of Ransomware Attacks, Global Cybersecurity Report Warns
A new Semperis report reveals AI is fueling 84% of ransomware attacks, with experts warning of autonomous threats and risks to corporate identity systems.
POLICY WIRE — Washington, USA — A comprehensive global cybersecurity study has revealed that artificial intelligence is now a primary driver in the digital criminal landscape, with 84% of ransomware victims reporting that attackers utilized AI to facilitate their operations. The 2026 Ransomware Risk Report, published by the firm Semperis, surveyed 1,200 IT and security professionals across eight industry sectors and nine countries to map the evolving threat.
The findings indicate that 44% of organizations surveyed experienced attacks that specifically targeted their own internal AI systems. Furthermore, 69% of victims reported suffering multiple successful breaches within the past year, while 59% of those polled admitted to paying a ransom at least once. Identity systems, which serve as the gatekeepers for network access, were targeted in 87% of the documented attacks.
Beyond the technical impact, the report highlights a severe human cost. Over one-third of victims faced physical or psychological threats, and 14% of organizations noted that staff members or their families were forced to relocate or change jobs due to the intensity of the harassment.
Chris Inglis, a former White House National Cyber Director and current Semperis board member, described AI as a significant inflection point in cyber warfare. He warned that the industry is already witnessing the rise of autonomous ransomware attacks, which allow even low-level actors to execute sophisticated breaches by leveraging AI to write malicious code and identify system vulnerabilities.
According to Inglis, AI expands the scale and scope of potential threats by democratizing access to advanced hacking tools. He noted that while the fundamental goal of criminals remains the same—extracting money by holding valuable data at risk—the technology allows them to tailor social engineering schemes and automate intrusions with unprecedented speed.
Tomer Bar, associate vice president of security research at Semperis, added that AI enables developers to produce more complex malicious code while reducing development time. This rapid evolution makes it increasingly difficult for law enforcement to attribute attacks, as attackers can now use large language models to test their code and eliminate errors in encryption algorithms.
The report also raises alarms regarding the integration of AI agents within corporate environments. Courtney Guss, associate vice president of crisis management at Semperis, cautioned that autonomous agents can become security liabilities if they seek excessive permissions. Guss described these agents as potential ticking time bombs that, if compromised, could grant attackers god mode access to sensitive infrastructure.
Despite these risks, Inglis cautioned against the assumption that AI inherently creates new vulnerabilities. He argued that the core issue remains a failure to control network access, and he urged organizations to pivot toward using AI for defensive purposes to counter the growing threat landscape.
Reporting by Policy-Wire (PW)
