Certification and Conformity Assessment

In accordance with the European Artificial Intelligence Regulation (AI Act) high-risk AI systems are subject to formal conformity assessment procedures before they are placed on the market or put into service. These procedures may take two main forms:

• The first is the provider’s internal conformity assessment. This generally applies to high-risk AI systems that are not embedded in products already covered by sector-specific European Union legislation.
  
  The provider is responsible for verifying internal compliance with the AI Act's requirements, including risk management, data governance, technical documentation, record-keeping, transparency, human oversight, and system robustness. The outcome of this process is formalised through an EU declaration of conformity.
• The second form is a third-party assessment by notified bodies. This applies when AI systems are incorporated into products that are regulated under sector-specific legislation, such as medical devices, machinery, or toys.
  
  In such cases, an independent designated body carries out the assessment to verify the system’s conformity before it is placed on the market. This process may result in the product receiving CE marking.

Notified bodies are independent entities that are accredited at a national level to carry out technical audits, review documentation, test systems, and verify that development processes comply with the requirements of the AI Act.

A compliant system will obtain a CE declaration of conformity, confirming that it meets the requirements of the Regulation. This certification will be mandatory for the lawful commercialisation of high-risk AI systems in the European Union.

Conformity assessment is therefore a central element of the AI Act’s governance framework for high-risk systems, ensuring that only compliant systems are made available and used within the European market.