54% of Enterprises Face AI Agent Security Incidents

A survey by VentureBeat Pulse Research highlights a concerning security gap in enterprise AI agent management. Over 54% of the 107 enterprises surveyed reported experiencing a security incident or near-miss involving AI agents. Despite this, many companies continue to allow agents to share credentials, increasing their vulnerability to breaches. The survey also found that most enterprises rely on security solutions from major providers like OpenAI and Google, yet plan to change their tooling within the year. This underscores the urgent need for improved security measures tailored to AI agents.