Healthcare AI Agents Face Security Scrutiny as Survey Finds 72% Unapproved Deployments

U.S. health systems are moving AI agents into clinical and administrative work faster than governance teams can formally approve or monitor them, creating a security problem around automated actions as well as data access. A survey of 250 U.S. healthcare identity-security and AI-strategy leaders found that 72% say AI tools or agents are deployed without formal IT approval, while 88% expect agents to operate with at least some autonomy.