At the Navigate 2026 conference in Austin, SailPoint Technologies highlighted the growing security risks posed by autonomous AI agents. Industry data indicates there are now 109 machine identities for every human user. Furthermore, a new AI agent is created every 4.5 seconds, while tasks on Amazon Bedrock AgentCore increased 15 times in the first half of 2026.
SailPoint is shifting its focus toward real-time identity governance and runtime enforcement to manage these non-human identities. The company recently acquired Entro Security, adding approximately 1,200 discovery sources for non-human identities to its platform. This integration, currently in internal testing, is scheduled for release to customers in November.
The company's research shows that while 80% of organizations believe their security tooling gap is manageable, only 15% can provision machine access in real time. SailPoint is addressing this through its Autonomous Identity approach, which includes just-in-time authorization and a Lineage Map to track access chains from humans to agents and data.
Practical challenges remain significant, as one Fortune 500 company discovered 10,000 agents on its network despite initial reports of none. SailPoint’s Agentic Accelerator program has enabled some Fortune 50 clients to move from deployment to outcomes in under a day, emphasizing the need for organizations to govern human and machine identities under a unified policy.
