Article 89: Monitoring actions

Article 89 gives the AI Office an operational mandate for monitoring actions over GPAI model providers. It is not about supervision of Union institutions.

1. Monitoring implementation. The AI Office may take the necessary actions to monitor effective implementation of and compliance with the Regulation by providers of general-purpose AI models, including their adherence to approved codes of practice (Article 56, not to be confused with the codes of conduct in Article 95). 2. Downstream provider complaint. Downstream providers may lodge a complaint alleging an infringement. It must be duly reasoned and state at least: the contact point of the model provider, the facts, the provisions concerned and the reasons, and any other relevant information.

• Article 88: the exclusive Commission and AI Office powers over Chapter V.
• Article 56: the GPAI codes of practice whose adherence may be monitored.
• Article 53: the baseline GPAI provider obligations that downstream complaints usually concern.
• Article 101: the GPAI-specific fine regime.
• Article 113: dates of application.

For downstream providers: build the complaint file with a contact point, the facts, the articles engaged and the evidence. A vague complaint does not meet Article 89(2).

For GPAI providers: document adherence to the code of practice and prepare for AI Office monitoring actions, which are distinct from national investigations into systems.

Article 89

Monitoring actions

1. For the purpose of carrying out the tasks assigned to it under this Section, the AI Office may take the necessary actions to monitor the effective implementation and compliance with this Regulation by providers of general-purpose AI models, including their adherence to approved codes of practice.

2. Downstream providers shall have the right to lodge a complaint alleging an infringement of this Regulation. A complaint shall be duly reasoned and indicate at least:

(a) the point of contact of the provider of the general-purpose AI model concerned;

(b) a description of the relevant facts, the provisions of this Regulation concerned, and the reason why the downstream provider considers that the provider of the general-purpose AI model concerned infringed this Regulation;

(c) any other information that the downstream provider that sent the request considers relevant, including, where appropriate, information gathered on its own initiative.