AI agent governance and runtime control are the real conference agenda

At a glance

What this is: This conference agenda is focused on AI agent governance, runtime control, and operational use cases, with sessions that treat agent identity and security oversight as the central design problem.

Why it matters: It matters because IAM, PAM, and AI security teams now have to govern agents that act like software actors with access, timing, and tool-use decisions that traditional approval flows do not handle well.

👉 Read C1.ai's Transform 2026 agenda for AI agent governance and security

Context

AI agent governance is emerging as a distinct security problem because the control question is no longer only what a model says, but what an agent can do, when it can do it, and under which runtime constraints. Once agents can chain tools, move across systems, and operate with delegated access, the governance boundary shifts from prompt safety to identity, privilege, and execution control.

This conference agenda reflects that shift across AI strategy, security operations, and infrastructure design. The identity angle is genuine: agents increasingly behave like non-human identities that need scoped access, auditability, and lifecycle controls, even when the broader discussion also touches runtime architecture and AI operations.

Key questions

Q: How should organisations govern external tools used by AI agents?

A: Organisations should review external tools as security inputs, not convenience features. Each tool needs ownership, approval, metadata inspection, and ongoing monitoring for hidden instructions or unexpected behaviour. If an AI agent can act on a tool, then the tool’s provenance and control status should be governed like any other sensitive integration.

Q: Why do AI agents create more risk than traditional automation?

A: AI agents create more risk because they can interpret context, choose actions, and invoke tools autonomously. Traditional automation follows fixed rules, but an agent can be manipulated into using its own authority in unintended ways. That makes permission scope, tool boundaries, and monitoring more important than model accuracy alone.

Q: What should security teams audit before allowing shared agents into production?

A: Audit how identity is propagated from the initiating user to the agent and then to every tool call, especially in shared channels. If you cannot show which user’s authority was exercised at each step, the agent is not ready for production access.

Q: How do organisations decide whether AI governance is strong enough for autonomous agents?

A: Organisations should ask whether their controls can observe API calls, tool use, and policy decisions outside the browser. If an agent can act without passing through the organisation’s visible control points, the governance model is incomplete. Autonomous workflows need explicit visibility into machine-to-machine execution paths.

Background and context

What the conference agenda says about AI governance maturity

The agenda shows that AI governance is moving from policy statements to operational control design. Sessions on visibility, framework selection, and judgment under AI pressure suggest that organisations are no longer satisfied with abstract principles alone. They want a playbook for ownership, runtime policy, escalation, and oversight. That is a sign of maturity, but it also exposes a gap: many programmes still lack a concrete model for how AI decisions map to existing IAM, PAM, and risk governance processes.

Practical implication: map AI governance responsibilities into existing identity and risk ownership structures before agent deployment expands.

NHI Mgmt Group analysis

Agent governance is now an identity problem disguised as an AI problem. Once agents can select tools, persist state, and act across systems, the meaningful question becomes who or what is authorised to do the work. That moves the control conversation into IAM, PAM, and NHI lifecycle management, not just model policy. Practitioners should treat agent identities as governed access subjects, not as incidental implementation detail.

Runtime choice now shapes the security boundary for agentic systems. A sandbox is only meaningful if it constrains access to secrets, networks, and downstream execution paths. If the runtime cannot limit those dimensions, the agent inherits more authority than teams usually realise. The practical takeaway is that infrastructure design and identity control now have to be reviewed together, especially where agents touch production workflows.

AI governance is becoming operational rather than aspirational. Conference agendas that include framework selection, judgment under pressure, and production agent deployment show that teams are moving past policy-only discussions. The next gap is not whether an organisation has principles, but whether those principles translate into access scope, auditability, and escalation rules. Practitioners should expect governance reviews to become more concrete and more technical.

Agentic AI will expose weak ownership models in existing identity programmes. Many IAM and PAM processes still assume a stable human user at the centre of access decisions. Agents break that assumption because they can appear and disappear dynamically, inherit privileges through orchestration, and execute at machine speed. The result is a control gap that requires explicit ownership, lifecycle, and review processes for non-human actors.

From our research:

• 92% of organisations expose NHIs to third parties, raising concerns about supply chain security, according to the Ultimate Guide to NHIs.
• 90% of IT leaders say properly managing NHIs is essential for a successful zero-trust implementation.
• Forward view: Read the Guide to the Secret Sprawl Challenge for the remediation patterns that matter when agent workflows inherit secrets and credentials.

What this signals

Agentic AI is forcing identity programmes to expand beyond human users and static service accounts. The practical issue is no longer whether an agent exists, but whether the organisation can assign ownership, revoke authority, and prove what it did. Teams that cannot model that boundary will end up with shadow AI behaviour inside otherwise mature access frameworks.

Governance teams should expect pressure to connect AI policy to access control evidence. That means moving from high-level principles to control statements that reference audit logs, scoped credentials, and offboarding paths. The organisations that do this well will be able to explain not just what the agent is allowed to do, but why the allowance can be trusted.

For practitioners

• Define agent identities explicitly Classify each agent, workflow bot, or AI assistant as a governed non-human identity with named ownership, scoped permissions, and review cadence. If an agent can call tools or touch data, it needs an access record as disciplined as any service account.
• Separate sandboxes from production authority Use isolated runtimes for experimentation and limit what production agents can reach through network policy, secret boundaries, and tightly scoped service credentials. The goal is to prevent a test runtime from quietly becoming a production access path.
• Tie agent actions to audit evidence Log the agent decision, the tool call, the credential used, and the resulting side effect so reviewers can reconstruct the chain of action. Without that evidence, incident review and governance sign-off both become guesswork.
• Map agent workflows to existing IAM and PAM controls Place agent onboarding, approval, privilege review, and offboarding into the same governance model used for high-risk accounts. That prevents AI projects from creating a parallel access system that bypasses standard control ownership.

Key takeaways

• AI agents create a governance problem because they act like non-human identities with dynamic access and decision-making authority.
• The most relevant control question is whether runtime boundaries, credentials, and audit evidence can keep agent behaviour inside approved scope.
• Security teams should map agent workflows into existing IAM and PAM ownership models before production use expands.

Key terms

• Agent Identity: An agent identity is the set of attributes, credentials and permissions assigned to an autonomous software entity. It is treated as a non-human identity because it can authenticate, act on systems and accumulate access over time, which creates governance, audit and lifecycle obligations similar to other production identities.
• Agent Runtime: The agent runtime is the execution environment where an AI agent reads data, calls tools, and carries out actions. It matters because the runtime is where identity, policy, and filesystem boundaries either hold or fail. If those boundaries are weak, the agent becomes a high-privilege path into the environment.
• Delegated Access: Delegated access is permission granted to one identity to act on behalf of another user, service, or system. In NHI environments, this usually appears in OAuth-connected apps and automation tooling. It is powerful, but it must be tightly scoped and reviewed because it can persist long after the original business need ends.
• Non-Human Identity (NHI): A digital identity assigned to a non-human entity such as a software application, service account, API key, bot, machine, or AI agent that enables it to authenticate and interact with systems without direct human involvement. NHIs now outnumber human identities in most enterprises by 25 to 50 times.

What to expect at the briefing

C1.ai's full article covers the conference agenda and speaker lineup this post intentionally leaves at a higher level:

• The full agenda and session sequencing across AI strategy, security, governance, and infrastructure topics.
• Named speakers and the specific themes each session is intended to cover at Transform 2026.
• On-site event details, including venue context and the live conference experience for attendees.

👉 C1.ai's full agenda provides the speaker lineup and session context behind the conference themes.

Deepen your knowledge

The NHI Foundation Level course, the industry's only accredited NHI security programme, covers NHI governance, machine identity security, and secrets management. It helps practitioners align identity control with the realities of agentic systems and other non-human actors.