According to ChannelE2E, a majority of companies—59 percent—are operating AI agents without implementing identity and access controls. The report highlights a significant security gap in enterprise AI deployments, as these systems operate outside standard identity management frameworks.