Skip to content
IAM Insights

Explainers · Glossary

Agent sprawl

agentic sprawl · genteck sprawl · AI agent sprawl

The uncontrolled growth in number of AI agents created by employees across platforms, many unused or unowned, which outpaces the organisation's ability to inventory and govern them.

In our briefings

The claim each of the newest briefings on Agent sprawl would have you leave with.

  1. Joy Chik, who leads Microsoft Entra, wants every AI agent made a first-class identity and governed with the same rigor as human identities: inventoried, with clear ownership and governed access. She says the same Zero Trust principles apply to agents and employees, and that the same tools can now manage both. As she describes Entra Agent ID, each agent receives its own identity, and requiring a human sponsor to govern that identity and its lifecycle "helps prevent orphaned agents". Onboarding and retiring agents can be automated, and Conditional Access policies can block risky agents and set least-privilege and just-in-time guardrails. The post does not say what retiring an agent revokes, or whether Agent ID was then in preview or generally available.

    Microsoft wants AI agents governed like employees, each with a human sponsorBI-FIND-2026-2404

  2. Singapore's IMDA offers agent identity practices for organisations to consider in the interim, and names the gaps that make them interim: today's authorisation systems typically use static scopes, and authentication systems, typically built around a single person, struggle when agents act for several users or spin up sub-agents. Each agent should have its own unique, cryptographically verifiable identity, tied to a supervising agent, a human user or a department. The capacity it acts in, such as on its own or for a named user, should be recorded, and every agent identity issued and tracked from a central system. Authorisations should generally be scoped, time- or session-bound, non-transferable and least-privilege, and, as a rule of thumb, a user should not be able to give an agent permissions beyond the user's own. IMDA points to OpenID's and the Cloud Security Alliance's agent identity papers for more.

    Singapore's regulator: each AI agent should have its own identity and, as a rule, no more access than its userBI-FIND-2026-0140

  3. Board-level guidance now names AI-agent identity outright. The 2026 fifth edition of the National Association of Corporate Directors (NACD) and Internet Security Alliance (ISA) Director's Handbook on Cyber-Risk Oversight asks directors: 'How are we managing AI identity (i.e., ensuring AI agents are only granted protected access to authorized data sources, and nothing else)? How is this tested for performance and security effectiveness?' Agent identity is now a director-level oversight question, not only a security-team topic.

    The board's own handbook now asks about AI agent identity by name, and the first answer is an inventoryBI-FIND-2026-0046

Cited in 4 briefings

Every briefing that discusses this term, newest first.

  1. Microsoft wants AI agents governed like employees, each with a human sponsor

    Article · · Microsoft Security Blog · Joy Chik · Jan 20, 2026

  2. Singapore's regulator: each AI agent should have its own identity and, as a rule, no more access than its user

    Article · · Infocomm Media Development Authority (Singapore) · May 20, 2026

  3. The board's own handbook now asks about AI agent identity by name, and the first answer is an inventory

    Podcast · · The Non-Human & AI Identity Podcast · NHI & AI Identity Summit : The CISO Imperative · Aug 28, 2026

  4. Rubrik ties agent recovery to protected backups and puts identity oversight in Agent Cloud

    Podcast · · AI Security Podcast · Why AI Agents Fail in Production: Governance, Trust & The "Undo" Button · Jan 23, 2026