The decision

Is the operating environment ready to control an AI-assisted workflow?

Readiness is not the same as enthusiasm, budget, or product availability. This assessment focuses on whether the organization can define acceptable use, protect data, integrate systems, train operators, and learn from incidents.

Input

Six current-state operating selections.

Output

A score, gap interpretation, and copyable action memo.

Owner

The program leader accountable for policy, controls, and pilot approval.

Evidence to prepare

Answer from records, not impressions.

  • Mission: service hours, critical decisions, response targets, and consequences of failure.
  • Policy: approved uses, prohibited uses, review owners, and change control.
  • Data: classification, access, retention, deletion, and audit evidence.
  • Integration: system inventory, interfaces, identity, telemetry, and degraded modes.
  • People: training records, coaching plans, and escalation competence.
  • Incidents: SOPs, drills, after-action records, and corrective-action closure.

Working tool

Score the current operating posture

Select the statement that best matches evidence available today, not the intended future state.

Interpretation

Use the weakest factor to shape pilot scope.

A composite score can hide one decisive weakness. Review the factor-level guidance before discussing a launch date.

Policy gap

Limit work to synthetic data and policy drafting until approved uses and owners exist.

Data gap

Do not introduce operational records until access, retention, and deletion are controlled.

Integration gap

Pilot an isolated workflow with manual handoff and measurable degraded operation.

Literacy gap

Train reviewers to detect unsupported claims, automation bias, and missing evidence.

Incident gap

Standardize the human process before adding an AI dependency.

High criticality

Require stronger testing, fallback, monitoring, and named approval authority.

Method and limits

A conversation structure, not a maturity certification.

The browser-local heuristic combines six ordinal selections. It is designed to expose discussion gaps and produce actions, not to certify conformance with NIST, ISO, law, contract, or an insurer's requirements.

Validate every action with legal, privacy, IT, HR, labor, records, and operational stakeholders appropriate to the workflow.

Review the NIST AI Risk Management Framework