The AI Identity Problem
Speaker: Paul Querna CTO, Co-founder ConductorOne
Main Takeaways
- ✅
Traditional identity systems were built for a manageable number of long-lived human identities.
- ✅
AI agents come in three forms: company agents, personal productivity agents, and SaaS-embedded agents.
- ✅
Company agents may scale 1–2x your employee count; personal agents could number in the hundreds per person per year.
- ✅
SaaS-based agents will grow with the number of applications and integrations a company uses.
- ✅
This leads to identity counts that can grow from thousands to over 100,000, even in small-to-midsize businesses.
- ✅
Organizations must rethink identity governance to support short-lived, task-based, and high-volume AI identities.
Learn More
blogThe State of Agentic AI in Identity Security
Security leaders aren’t just experimenting with AI—they’re rapidly integrating agentic AI into high-stakes workflows like identity provisioning and network defense. ConductorOne’s 2025 Identity Security Outlook Report reveals that urgency is driving adoption, but with thoughtful strategies and real concern for risk.
blogThe AI Identity Problem
AI agents are changing the identity landscape, but legacy IAM systems weren’t built for them. These fast-moving, ephemeral identities require dynamic access, machine-speed decisions, and a new model of governance. In this post, we break down the AI identity problem—and what it takes to solve it.
blogThe Divergent AI Landscape: Making Sense of Agentic AI
Explore the evolving world of agentic AI. Learn the three core types of AI agents—company, employee, and agent-to-agent—and the identity, security, and governance challenges they introduce.
