Interim Measures under AI Act

Legal Framework

Interim measures under the AI Act are grounded in the market surveillance architecture established by Regulation (EU) 2019/1020, as incorporated into the AI Act's enforcement regime. Article 85 of the AI Act provides that any natural or legal person may lodge a complaint with the relevant market surveillance authority where there are grounds to believe an infringement has occurred. These complaints feed directly into market surveillance activities and must be handled under the dedicated procedures established by national authorities.

The competence to take interim protective action rests primarily with national market surveillance authorities, except where an AI system is based on a general-purpose AI model provided by the same provider—in which case the AI Office acts as the market surveillance authority at Union level, as clarified by Recital 161. Recital 170 confirms that the complaint mechanism operates without prejudice to existing judicial and administrative remedies under Union and national law, ensuring that interim measures form one layer within a broader enforcement ecosystem.

The substantive basis for interim measures arises where a high-risk AI system presents an immediate and substantiated risk to fundamental rights, safety, or public security. Authorities may restrict or prohibit the system's availability on the market, require its withdrawal, or impose conditions on its use pending the provider's corrective action or a full compliance assessment.

Key Developments

The AI Act entered into force on 1 August 2024, with enforcement staged through 2026 and 2027. No case law or enforcement decisions on interim measures have yet emerged, as the relevant provisions are not yet fully applicable. However, the framework draws directly on the established practice under Regulation (EU) 2019/1020, where market surveillance authorities have exercised analogous powers to recall products, impose sales bans, and require corrective action for non-compliant goods posing immediate risks. Member States are currently designating their national market surveillance authorities and establishing the procedural infrastructure to receive and act on complaints under Article 85.

Practical Guidance

• Map your market surveillance authority: Identify the competent national authority for each Member State where your high-risk AI system is placed on the market, and the AI Office's role where general-purpose AI models are involved under Recital 161.

• Establish a complaint intake protocol: Article 85 allows any natural or legal person to lodge complaints. Maintain internal procedures to monitor and respond to complaints escalated by authorities, as these trigger market surveillance activity.

• Prepare for provisional restrictions: Develop contingency plans for scenarios where authorities impose interim measures—product withdrawal, market restrictions, or use conditions—pending compliance verification.

• Document risk mitigation measures: Maintain real-time evidence of safety controls, conformity assessments, and corrective action capacity to demonstrate readiness if an authority initiates interim proceedings.

• Coordinate across remedies: Recital 170 preserves existing judicial and administrative remedies. Ensure legal strategy accounts for parallel proceedings before national courts, data protection authorities, and AI-specific market surveillance bodies.