Join our Newsletter — 33% off our NHI Course
Agentic AI & Autonomous Identity

Elicitation

← Back to Glossary
By NHI Mgmt Group Updated October 30, 2026 Domain: Agentic AI & Autonomous Identity

Elicitation is the MCP mechanism that pauses a task and asks a person for information or confirmation. It is the protocol’s only built-in human checkpoint, but it still requires external controls to ensure the responder is the same person who started the flow and that the answer is recorded accurately.

What Elicitation Means in MCP

Elicitation is the Model Context Protocol step that interrupts a task to ask a person for information or approval. It is designed as a human checkpoint, not a trust guarantee. The protocol can stop and ask, but it cannot by itself prove who answered, whether the response was genuine, or whether the record is complete.

Why Elicitation Exists

In practice, elicitation is the point where an automated flow hands control back to a human for a decision the system should not guess. That is useful when the task needs confirmation, missing input, or explicit consent. It reduces blind automation, but it also introduces a dependency on the surrounding application to manage who is being asked and how the answer is bound to the session.

How Elicitation Should Be Interpreted

The important point is that elicitation is a protocol interaction, not an identity check. A prompt on its own does not establish that the same person who started the workflow is the one responding. It also does not ensure the answer is captured in a durable, attributable, and tamper-evident way. Those assurances come from the host application, authentication layer, audit trail, and session management around the MCP flow.

Where Elicitation Fits in Secure MCP Design

Elicitation should be treated as one control point inside a broader trust boundary. It is most valuable when the system needs to pause before a sensitive action, such as disclosing data, invoking a tool, or continuing a workflow with side effects. The design question is not only whether a prompt appears, but whether the surrounding control plane can bind the response to the right user, preserve the decision, and prevent replay or substitution.

Risk and Threat Considerations

Elicitation creates a human-in-the-loop moment that can be abused if the surrounding controls are weak. An attacker who can hijack a session, confuse the user, or manipulate the prompt can turn the checkpoint into a false sense of safety rather than a real guardrail.

Failure mechanism: The task pauses, but the responder may not be the original user, the response may be misattributed, or a malicious prompt may steer the person into approving an action they would otherwise reject.

Impact: Sensitive actions can proceed under incorrect authority, and the organisation may lose confidence in the integrity of approvals, audit evidence, and downstream task execution.

Standards & Framework Alignment

This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.

NIST SP 800-53 Rev 5 and NIST CSF 2.0 set the governance and control requirements practitioners need to meet.

FrameworkControl / ReferenceRelevance
NIST SP 800-53 Rev 5IA-2 — Identification and Authentication (Organizational Users)Elicitation depends on knowing which authenticated user is responding.
AU-2 — Event LoggingElicitation decisions need auditable records of who answered and what was approved.
AC-6 — Least PrivilegeElicitation often gates sensitive actions that should only be available with minimal authority.
Recommendation — Bind elicited responses to an authenticated user before continuing the workflow. Log the elicitation prompt, responder identity, timestamp, and resulting action. Limit the actions that can be triggered after an elicited response.
NIST CSF 2.0PR.AA-05 — Identity Management, Authentication and Access ControlElicitation sits inside identity-aware access decisions that must be verified.
DE.CM-01 — Network and System Activity Is MonitoredElicitation flows benefit from monitoring for abnormal prompts, approvals, and session abuse.
Recommendation — Require strong identity checks before accepting a human response as authoritative. Monitor elicitation activity for unusual approval patterns and session anomalies.

Practitioner Guidance

Why practitioners should care: Elicitation is only as strong as the controls wrapped around it. Treat it as a workflow interruption that needs session binding, user verification, and reliable logging, not as proof of intent or identity.

Common misunderstanding: Teams sometimes assume that a human prompt automatically makes the flow safe. In reality, the main failure is often not the prompt itself, but the gap between the prompt and the authenticated user, the recorded decision, and the action that follows.

Practitioner takeaway: Use elicitation to slow down high-impact steps, then verify that the application can attribute the response to the right actor and preserve that decision for review.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Reviewed and updated by the NHIMG editorial team on October 30, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org