AI agent governance dominates ISC2 Security Congress 2026 agenda

At a glance

What this is: This is a conference planning guide that distils ISC2 Security Congress 2026 into role-based session picks, with AI agent governance as the dominant thread.

Why it matters: It matters because identity teams now need to treat agent access, accountability, and governance drift as agenda-level priorities rather than side topics.

By the numbers:

• Nearly 200 sessions run across five days at ISC2 Security Congress 2026.

👉 Read Clarity Security’s guide to ISC2 Security Congress 2026 session planning

Context

ISC2 Security Congress 2026 is a conference planning problem as much as a content problem: the agenda is broad, but the identity governance signal is concentrated in AI agents, access control, and operational accountability. For practitioners, the question is which sessions help you govern who or what can act, not just which talks are popular.

The guide groups sessions for Security Leaders, IAM teams, GRC and risk, and Security Architects and Engineers. That structure is useful because the same underlying issue appears in different forms across the agenda, from governance drift and third-party risk to AI context leaks and least-privilege design.

Key questions

Q: How should security teams prepare for AI security governance at large cloud conferences and enterprise events?

A: Security teams should use these events to pressure test how they govern AI interactions, not just market their AI plans. Priorities include access scoping, secrets hygiene, logging, and approval workflows for autonomous actions. Teams should leave with a clear view of who can deploy, change, or query AI systems, and how those actions are reviewed and contained.

Q: Why do AI agents create a governance problem even when the protocol is approved?

A: Because the protocol only moves traffic, it does not define authority. An approved MCP, OAuth, SDK, or browser path can still carry an agent that has no clear owner, no bounded intent, or excessive privilege. Governance fails when organisations confuse transport approval with identity approval.

Q: What are the signs that AI governance is failing in the enterprise?

A: Common warning signs include rapid growth in AI use without matching policy coverage, sensitive files being copied into personal accounts, and a large share of AI apps carrying high or critical risk. Another indicator is weak visibility into who is using which tools and what data they are sending. If teams cannot answer those questions, governance is not working as intended.

Q: What should teams do when an AI workflow starts using more data or tools than expected?

A: Treat that as a new governance event, not a minor configuration change. Pause the expansion, document the new scope, decide whether the original approval still applies, and confirm what evidence is needed before the workflow continues. If the change cannot be justified in writing, the workflow has outgrown its authorised boundary.

Background and context

Why AI agent governance is now an identity problem

AI agents become an identity problem when they can request access, use tools, and act on context without the controls built for human-paced workflows. The core issue is not just model quality. It is delegated authority, delegated context, and delegated action occurring faster than manual review cycles can observe. That changes how teams think about approval, evidence, and blast radius. In practical terms, the identity model has to cover what the agent may do, which tools it may use, and when that authority expires. See the AI Agent Authorisation Guide. See the AI Agent Authorisation Guide.

Practical implication: Treat AI agent access as a governed identity lifecycle, not as a simple application integration.

How governance drift shows up in agentic environments

Governance drift happens when approved use cases expand into new data, new tools, or new decisions without a corresponding update to policy and control evidence. In agentic systems, that drift is often invisible because the workflow still appears legitimate on the surface while the scope of access keeps widening underneath. This is why tabletop exercises, conditional approvals, and runtime controls matter more than annual sign-off alone. The issue is not merely whether an agent exists, but whether its permissions and context remain aligned with the original approval boundary. See the Agentic AI Identity Guide. See the Agentic AI Identity Guide.

Practical implication: Define explicit scope boundaries for each agent and review any expansion as a new governance event.

Why context leaks are a control failure, not just a data issue

Context leakage in MCP-style or agent-integrated environments is a control failure because the agent can expose or reuse sensitive information across prompts, tools, and sessions. Once context becomes a transport for decisions, it also becomes a control plane for privilege and disclosure. That creates a new risk class where sensitive input, tool access, and downstream action are coupled. Security teams should treat context boundaries as enforceable policy boundaries, not as a user experience detail. See the MCP Security Guide. See the MCP Security Guide.

Practical implication: Enforce data handling and tool-use policy at the context boundary, not after the agent has already acted.

NHI Mgmt Group analysis

AI agent governance is now an identity discipline, not an app-layer add-on. The agenda in this conference guide makes that clear: sessions on governance, context leaks, admissibility, and least privilege all point to the same operating reality. When agents can act across tools and data sources, identity is the control surface that determines whether the system stays bounded or drifts into unchecked authority. Practitioners should read this as a shift from tool adoption to governed execution.

Governance drift is the more common failure mode than outright compromise. A security programme can approve an AI use case and still lose control when the same workflow expands into new data, new decisions, or new integrations. That is why lifecycle governance has to track scope changes, not just initial approvals. The risk is not only malicious abuse but quiet boundary expansion that leaves policy behind. Agentic AI Identity Guide is relevant because the problem is identity scope, not model novelty.

The new control gap is the gap between granted authority and reviewed authority. Traditional review cadences assume privileges persist long enough to be inspected and recertified. In agentic environments, access can be created, used, and expanded inside the same operational window, which leaves nothing stable for a periodic review to catch. Practitioners should treat that as an assumption failure in current governance models, not just a tooling gap.

Conference agendas are becoming governance maps for the identity stack. The sessions highlighted here span human, NHI, and AI agent concerns because the operational question is converging: who or what is allowed to act, under what evidence, and with which accountability trail. That convergence is important because identity programmes that stay siloed will miss the points where delegation, context, and authorisation intersect. The implication is broader programme alignment, not a standalone AI policy.

Shadow AI is an NHI problem before it becomes an AI problem. The guide’s emphasis on hidden agents and invisible attack surfaces shows that unmanaged agents behave like unmanaged non-human identities: they exist, they act, and they accumulate privilege faster than owners can enumerate them. That is why inventory, ownership, and offboarding matter as much for AI agents as they do for service accounts. Security teams should govern agent inventories with the same seriousness they apply to other non-human estates.

From our research library:

• 19% of organisations give AI systems dramatically more access than human employees, nearly one in five granting unrestricted privilege, according to the 2026 Infrastructure Identity Survey.
• 70% of organisations grant AI systems more access than they would give a human employee performing the exact same job, according to the 2026 Infrastructure Identity Survey.
• Read next: Agentic AI Identity Maturity Model

What this signals

AI governance will increasingly be decided through identity controls, not policy statements. As more teams move from experimentation to production, the practical question becomes whether access can be bounded, reviewed, and revoked at the pace agents operate. That pushes programmes toward task-scoped authority and away from broad, persistent access.

Governance programmes that assume stable, reviewable privileges will misread agent behaviour. Agentic systems can request, use, and release access inside a single operational cycle, which means periodic reviews may never see the risky state at all. The programme implication is clear: review cadence alone is no longer enough.

Shadow AI will force identity teams to own discovery as well as policy. Once hidden agents appear in business workflows, the first control question is whether the organisation can inventory them, assign ownership, and retire them cleanly. That is the same lifecycle problem identity teams already know, but applied to systems that move faster than humans can track.

For practitioners

• Map conference sessions to governance decisions, not job titles. Build your Security Congress plan around the control questions you need answered, such as agent scope, approval boundaries, audit evidence, and incident attribution.
• Prioritise sessions on AI agent authority and delegated access. Focus on talks that address what agents may access, how permissions expire, and how human approval fits into the workflow.
• Use tabletop scenarios to test governance drift. Walk through a case where an approved AI workflow expands into new data or tools and decide who must approve the change, what evidence is required, and when to halt execution.
• Separate context leakage from model failure in your planning. Track which sessions address prompt injection, tool misuse, and insecure integrations so you can distinguish content-risk issues from identity-and-access issues.
• Treat shadow AI as part of your non-human inventory. Update discovery and ownership processes so unmanaged agents are found, assigned, and retired with the same discipline used for other non-human identities.

Key takeaways

• AI agents are now a governance and identity issue, not just a technology trend, because they can expand authority faster than review cycles can keep up.
• The conference agenda reflects a broader shift toward scope control, accountability, and context governance across production AI systems.
• Security teams should use events like this to validate their boundaries for agent access, ownership, and offboarding before those controls are tested in production.

Related NHI Mgmt Group resources: AI Agent Authorisation Guide · Agentic AI Identity Guide · Zero Trust for AI Agents · MCP Security Guide

Key terms

• AI Agent Governance: AI Agent Governance is the set of policies, controls, and oversight practices used to direct how autonomous software agents behave. It defines allowed actions, approval paths, identity boundaries, logging, monitoring, and accountability so agent decisions remain traceable, constrained, and aligned with business, security, legal, and ethical requirements.
• Governance Evidence Drift: The gap that appears when written governance controls lag behind the actual state of AI access, ownership, and activity. It is a practical audit problem because a policy can still exist while permissions, data exposure, or accountability have already changed underneath it.
• Shadow AI: AI agents, copilots, or connected tools operating without full visibility or governance from security teams. Shadow AI becomes an identity problem when those systems authenticate with unmanaged tokens, service accounts, or OAuth apps that can reach production resources.
• Delegated Authority Model: A delegated authority model defines who is allowed to approve, review, or execute control-related decisions across the enterprise. It helps ensure requests reach the correct responsible party, especially when control owners, managers, and process owners sit in different teams, regions, or systems.

What to expect at the briefing

Clarity Security’s full conference guide covers the operational detail this post intentionally leaves for the source:

• Role-by-role session picks for Security Leaders, IAM, GRC, and Architects so you can build a personalised schedule
• Session timing and conference logistics, including workshop windows, expo hall hours, and limited-seat tabletop sessions
• Specific talk descriptions and speaker pairings for governance, agentic AI, IAM, and Zero Trust tracks
• On-site planning cues such as booth location, follow-up opportunities, and networking events

👉 The full Clarity Security guide covers the role-based session map, on-site logistics, and the AI governance talks to prioritise.

Deepen your knowledge

NHI governance, agentic AI identity, and machine identity security are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are building or maturing an identity security programme, it is worth exploring.