The AI System Operator / End User is the person who uses an AI system to do a job. They are not the governing body, the builder or the risk owner. Their duty is to use the system only for the signed intended purpose, follow the oversight model Role 4.1 designed, refuse off-purpose use, and raise Act and incident signals when the system or the work would leave that purpose.

Standing duties

• Use only the authorised configuration for the authorised purpose. A convenient extra question to the model is often ACT02.
• Apply the oversight instructions for this system: when to accept, when to overrule, when to stop, when to escalate.
• Record interventions and anomalies in the path Role 4.1 and OPR02 (owned by Role 22.1) specified — not in a private chat.
• Refuse to operate if the roster, hook or explanation required by the oversight model is missing.
• Do not treat “I used it and nothing bad happened” as residual-risk acceptance or as validation.

Accountable to: Service owner / line manager of the business process (oversight design: Role 4.1)

Primary domain: Operate (use within signed purpose). Does not Direct, design, or accept residual risk.

Linked EN 18286 clauses: 4.1, 5.1, 7.4, 9.2, 9.6.2.1

Format: 3 hours of training, EUR 200.