Cyberattacks increasingly caused by unchecked AI agents

Infosecurity Magazine reports that the lack of strategy on the use and governance of AI agents is increasing organizations' risk of experiencing cybersecurity incidents.

According to research conducted by Cloud Security Alliance and Token Security, 65% of enterprises have experienced at least one AI agent-related cybersecurity incident within the past year. Over half of the respondents reported confidence in AI agent visibility in their network, but 82% have claimed finding undiscovered agents in the past 12 months. Researchers said forgotten agents that are not decommissioned properly could pose a cybersecurity risk, and incidents are affecting core enterprise operations. CSA encourages organizations to determine strategies for maintaining AI agent visibility, applying consistent lifecycle governance, setting operational boundaries for agents, and guiding decision-making with human approval.

"As agents gain greater autonomy, governance must evolve into a more unified, operational model that can sustain control at scale," said CSA Assistant Vice President of Research Hillary Baron.