Methodology Services About Blog Contact
← Back to Blog
EU AI Act January 7, 2025

EU AI Act: Authorised Representative, all you need to know

The EU AI Act introduces a wide range of obligations for providers of AI systems operating in the European market. For providers established outside the European Union, one of the most important requirements is the appointment of an Authorised Representative (AR). This article explains what an AR is, when one is required, what responsibilities they carry, and how to choose the right partner for this critical role.

What is an Authorised Representative?

An Authorised Representative is a natural or legal person located within the European Union who is formally mandated by a non-EU AI provider to act on their behalf with respect to the obligations set out in the EU AI Act. The AR serves as the provider's official presence in the EU, ensuring that regulatory requirements are fulfilled and that there is a responsible point of contact for authorities.

The concept mirrors similar requirements under other EU regulations, including the GDPR's Article 27 representative and the EU's product safety directives. However, the AI Act's AR role carries broader and more demanding responsibilities, reflecting the complexity and potential impact of AI systems.

When is an Authorised Representative required?

The obligation to appoint an AR arises in two main scenarios under the EU AI Act:

  • High-risk AI system providers outside the EU: Any provider of a high-risk AI system that is not established in the European Union must appoint an Authorised Representative before placing that system on the EU market or putting it into service within the EU. High-risk AI systems include those used in areas such as biometric identification, critical infrastructure management, education, employment, access to essential services, law enforcement, and migration management.
  • General-purpose AI (GPAI) model providers outside the EU: Providers of GPAI models, including foundation models and large language models, who make their models available in the EU market and are not established in the Union must also appoint an AR. This applies regardless of whether the model is provided directly to end users or to downstream developers who integrate it into their own systems.

In both cases, the AR must be established in one of the EU Member States and must accept the mandate in writing before the provider can lawfully operate in the European market.

Key responsibilities under Article 22

Article 22 of the EU AI Act sets out the core responsibilities of the Authorised Representative. These are not ceremonial duties; they carry real legal weight and practical significance.

Verification and record-keeping

The AR must verify that the provider has drawn up the required technical documentation and has carried out or had carried out the appropriate conformity assessment procedure. The AR must keep a copy of the technical documentation, the EU declaration of conformity, and any relevant documentation at the disposal of national competent authorities and market surveillance authorities for a period of at least 10 years after the AI system has been placed on the market.

This is not merely a filing obligation. The AR must understand the documentation well enough to respond to regulatory inquiries and demonstrate that the provider's AI system complies with the Act's requirements.

Regulatory liaison

The Authorised Representative serves as the primary point of contact for all interactions with EU authorities regarding the AI system. This includes national competent authorities in each Member State, the European AI Office, and market surveillance authorities. The AR must cooperate with these bodies on any action they take in relation to the AI system, including inspections, investigations, and requests for information.

Effective regulatory liaison requires not only legal expertise but also a thorough understanding of the AI system's technical characteristics, intended use, and risk profile. The AR must be able to communicate credibly with regulators who may raise complex technical and legal questions.

Risk mitigation

If the AR has reason to believe that an AI system presents a risk to the health, safety, or fundamental rights of persons, they must immediately inform the provider and recommend appropriate corrective measures. If the provider fails to take adequate action, the AR must take their own steps, which may include informing the relevant authorities or, in serious cases, withdrawing the system from the market.

This obligation means the AR cannot be a passive intermediary. They must actively monitor the AI system's compliance status and be prepared to act independently when the situation demands it.

Termination obligations

If the mandate between the AR and the provider is terminated, the AR's responsibilities do not vanish overnight. The AR must ensure an orderly transition, which may include transferring documentation to a successor AR or to the provider, and they may continue to bear certain obligations during a transitional period. This underscores the importance of establishing clear contractual terms from the outset.

Choosing the right Authorised Representative

Selecting an AR is one of the most consequential decisions a non-EU AI provider will make under the AI Act. The right choice can smooth your path to compliance; the wrong one can create regulatory exposure and operational friction. Three criteria are paramount:

Deep understanding of the EU AI Act

Your AR must have genuine expertise in the EU AI Act and the broader European regulatory landscape. This is not a role that can be filled by a general legal counsel or a compliance administrator without specialized knowledge. The AI Act is technically complex, and its requirements interact with other EU regulations (GDPR, product safety directives, sector-specific rules) in ways that require nuanced understanding.

Experience with AI systems

Regulatory expertise alone is not sufficient. The AR must understand how AI systems work: how they are trained, how they make decisions, what can go wrong, and how risks manifest in practice. Without this technical grounding, the AR will struggle to verify documentation, communicate with regulators, or assess risk effectively.

Operational capability

The AR role is not a one-time appointment; it is an ongoing operational commitment. Your AR must have the capacity to manage documentation, respond to regulatory inquiries promptly, monitor compliance, and act decisively when issues arise. They must also be reachable and responsive, since regulators may make time-sensitive requests. Evaluate potential ARs on their infrastructure, team, and track record, not just their credentials.

Why this role is critical

The Authorised Representative is more than a compliance formality. They are your organization's gateway to the European AI market. A competent AR protects your market access, shields you from penalties, provides early warning of regulatory developments, and demonstrates to customers and partners that you are committed to responsible AI.

Conversely, an inadequate AR, or the absence of one, puts your entire EU business at risk. Non-compliance can result in fines of up to EUR 15 million or 3% of global turnover, as well as orders to withdraw your AI system from the market.

How we can help

At Systems Impact, we combine deep expertise in the EU AI Act with hands-on experience in AI systems and governance. We help non-EU AI providers appoint qualified Authorised Representatives, prepare the required technical documentation, and build the compliance processes needed for sustained market access in Europe.

If you are looking for an Authorised Representative or need guidance on your EU AI Act obligations, get in touch. We are here to help you navigate this critical requirement with confidence.