Webinar Description
Key Takeaways
- Examines the security implications of AI agents acting as autonomous users within enterprise systems
- Addresses access sprawl, governance gaps and hidden operational costs associated with AI-driven workflows
- Designed for IT leaders, security professionals and operations managers overseeing AI adoption
- Provides a practical framework for establishing visibility and control over AI access
- Hosted by LastPass with strategic guidance from CTO Jason Rasmussen
Introduction
As organisations accelerate their adoption of artificial intelligence, a new category of security challenge has emerged: AI agents that operate as autonomous users within business systems. These agents interact with applications, access sensitive data and execute tasks on behalf of employees, often without the same oversight applied to human users. The webinar “Who’s Minding the Agents? Securing Access When AI Is the User” addresses this growing concern, offering IT and security leaders a structured approach to understanding and governing AI-driven access before it creates unmanageable risk.
About This Event
Hosted by LastPass, this 45-minute virtual session is led by Jason Rasmussen, the company’s Chief Technology Officer. The webinar focuses on the practical realities of AI agents operating within enterprise environments, examining how these automated systems create what Rasmussen describes as a new form of access sprawl. Unlike traditional user access management, AI agents often proliferate across tools and systems without centralised visibility, making it difficult for security teams to understand what data is being accessed and by which automated processes.
The Challenge of AI Agents as System Users
Traditional identity and access management frameworks were designed with human users in mind. Authentication flows, access reviews and audit trails assume a person is requesting and using credentials. AI agents disrupt this model fundamentally. They may be granted broad permissions to function effectively, yet operate continuously without the natural pauses that prompt human users to log out or request access renewals. This creates persistent access that can go unmonitored for extended periods.
The webinar explores how this dynamic leads to access sprawl—a condition where the number of connections between AI agents and business applications grows faster than security teams can track. Each new AI tool introduced by a department may establish its own integrations, credentials and data access patterns. Without governance mechanisms designed specifically for non-human identities, organisations risk accumulating technical debt alongside genuine security vulnerabilities.
Hidden Costs and Compliance Implications
Beyond direct security risks, ungoverned AI access carries operational and financial costs that often remain invisible until they compound into larger problems. Redundant tool subscriptions, duplicated integrations and inconsistent access policies across departments all contribute to inefficiency. From a compliance perspective, organisations subject to data protection regulations may struggle to demonstrate adequate controls over how AI systems process personal or sensitive information.
The session addresses these concerns by outlining a governance framework that brings AI agent activity under the same scrutiny applied to human users. This includes establishing visibility into which agents exist, what permissions they hold and how frequently they access critical systems.
Who Should Attend
The webinar is designed for senior technology and security professionals responsible for enterprise access policies. CTOs, CIOs and CISOs will find strategic value in understanding how AI adoption intersects with existing identity management programmes. Security engineers and IT managers tasked with implementing controls will benefit from the practical framework presented. Operations leaders concerned with cost management and process efficiency may also find relevant insights, particularly regarding the hidden expenses associated with uncoordinated AI tool deployment.
Establishing Governance Before Complexity Escalates
The central argument of the session is that organisations should establish AI access governance proactively rather than reactively. As AI agents become more capable and more deeply integrated into business workflows, the difficulty of retrofitting controls increases substantially. By implementing visibility mechanisms and access policies now, security teams can maintain oversight as AI adoption scales, reducing the likelihood of data breaches, compliance failures and operational inefficiencies that stem from ungoverned automation.