Webinar
Your Agents Have Access. But Do You Have Control?
New EMA research reveals a widening gap between deployed agentic AI and its governance.
Nearly every enterprise believes its AI agents have the right level of access — but only 33% consistently enforce least privilege. Join experts from Enterprise Management Associates (EMA) and Cequence Security as they unpack new research into the risks emerging as agentic AI moves into production. Learn where governance is falling short, why existing controls aren’t enough, and how security leaders can better govern every action an AI agent takes.
Confidence Is High. Enforcement Isn’t.
AI agents have moved in growing numbers beyond experimentation and into business-critical workflows. Yet new research from Enterprise Management Associates (EMA) shows that governance practices have not kept pace.
While 94% of enterprise IT and security leaders are confident their AI agents are not over-provisioned, only 33% actually give agents least-privilege access. Broad standing permissions, delayed authorization checks, and forgotten credentials are creating risks that many organizations cannot see until after something goes wrong.
In this webinar, we’ll examine what that gap looks like in production – and what security teams can do to close it.
What You’ll Learn
Why standing permissions leave AI agents with access that can outlive their original task
Why authentication/authorization are important, but controlling real-time AI agent behavior and actions is the key to scale and risk reduction
How abandoned pilots and unmanaged credentials create hidden exposure
Practical steps for governing AI agents across their full lifecycle
How to balance rapid AI adoption with effective governance and security
Featured Speakers
Christopher M. Steffen
CISSP, CISA, VP of Research at EMA
Randolph Barr
CISO at Cequence Security
About the Research
EMA surveyed 202 enterprise IT and security leaders – including CIOs, CTOs, CISOs, and IT directors – at organizations with at least 1,000 employees. Respondents represent organizations across North America and EMEA that are deploying or evaluating agentic AI in industries including technology, financial services, healthcare, and manufacturing.
Don’t Let Confidence Become a Blind Spot
Discover what AI agents are actually doing, where existing governance practices fall short, and how to enforce control when it matters most: at the moment an agent acts.
