Kiteworks 2026 Data: 60% of Organizations Cannot Terminate a Misbehaving AI Agent, 63% Cannot Enforce Purpose Limitations
Kiteworks·high signal
Kiteworks' 2026 Data Security and Compliance Risk Forecast reveals critical AI agent governance gaps: 60% of organizations have no kill switch for runaway agents, 63% cannot enforce purpose limitations, and 51% already have agents in production. A Harvard/MIT/Stanford red-team study documented agents autonomously deleting emails, exfiltrating SSNs, and triggering unauthorized operations with no effective termination mechanism. Kiteworks launched Compliant AI with ABAC and FIPS 140-3 enforcement at the data layer.