Webinar Description
Key Takeaways
- Explores just-in-time access strategies for developer environments where AI agents and non-human identities now outnumber human users
- Addresses the security risks of standing privileges in CI/CD pipelines, service accounts and autonomous AI workflows
- Covers run-time authorisation, automated access revocation and the transition from legacy access controls to JIT frameworks
- Relevant to CISOs, security directors, developers and PAM/IAM teams managing complex identity ecosystems
- Hosted by Delinea with participation from StrongDM
Introduction
Just-in-Time (JIT) Access for Developer Environments in the Era of Frontier AI Models – EMEA is a virtual webinar examining how organisations can secure access in environments where autonomous AI models, service accounts and CI/CD pipelines have fundamentally changed the identity landscape. Aimed at security leaders, developers and identity management professionals across the EMEA region, the session addresses a pressing operational challenge: standing privileges—permanent access rights that remain active regardless of whether they are needed—create expanding attack surfaces as non-human identities proliferate.
The timing reflects broader industry shifts. As frontier AI models gain autonomy and developer ecosystems grow more complex, traditional access control models struggle to keep pace. Every identity, whether human or machine, represents a potential entry point for attackers, making the elimination of unnecessary persistent access a security imperative rather than an operational nicety.
The Case Against Standing Privileges
Standing privileges have long been a convenience in enterprise environments, allowing administrators and automated processes to maintain continuous access to critical systems. However, this model assumes a relatively static identity environment with predictable access patterns. Modern developer ecosystems invalidate that assumption.
Contemporary software delivery relies on extensive automation. CI/CD pipelines execute deployments, service accounts interact with cloud infrastructure, and AI agents increasingly perform tasks that previously required human intervention. Each of these non-human identities requires credentials, and in many organisations, these credentials carry persistent privileges that remain active around the clock. The result is a dramatically expanded attack surface where compromised credentials—whether belonging to a developer or an automated process—can provide adversaries with sustained access to sensitive systems.
JIT access offers an alternative approach. Rather than maintaining permanent permissions, access is granted only when needed and revoked immediately after use. This model reduces the window of opportunity for attackers while maintaining the operational flexibility that engineering teams require.
Run-Time Authorisation and Automated Access Management
The webinar explores run-time authorisation as a practical mechanism for implementing JIT access at scale. Unlike traditional access controls that evaluate permissions at login or during periodic reviews, run-time authorisation assesses access requests in real time, considering contextual factors before granting temporary credentials.
Automation plays a central role in making this approach viable. Manual approval workflows cannot keep pace with the speed of modern software delivery, where deployments may occur dozens of times daily. Automated access granting and revocation enables organisations to enforce strict access policies without introducing friction that slows engineering velocity. The challenge lies in balancing security requirements with the operational demands of fast-moving development teams—a tension the session directly addresses.
Managing Human and Non-Human Identities
A significant portion of the discussion focuses on the growing complexity of identity management in DevSecOps environments. Non-human identities—including AI agents, service accounts and automated pipelines—now frequently outnumber human users in enterprise environments. Traditional Privileged Access Management and Identity and Access Management frameworks were designed primarily with human users in mind, creating gaps when applied to machine identities.
The webinar examines how organisations can extend JIT principles to encompass this broader identity ecosystem. This includes securing AI workflows where autonomous agents require access to production systems, databases and APIs. As AI models become more capable and are granted greater autonomy, the access management strategies governing their behaviour become increasingly consequential.
Who Should Attend
The session is designed for security and technology leaders responsible for access management strategy, including CISOs, CTOs and Directors of Information Security. Practitioners working directly with PAM and IAM implementations will find practical relevance, as will developers operating within environments where security and velocity must coexist. Organisations already investing in DevSecOps practices or deploying AI agents in production workflows represent the primary audience.
Event Format
Hosted by Delinea with participation from StrongDM, the webinar takes a technical and practitioner-focused approach. The format combines educational content with insights drawn from real-world implementation experience, offering attendees actionable perspectives on transitioning from legacy access models to JIT frameworks suited to the demands of AI-era development environments.