As the first comprehensive legislative attempt to regulate AI technologies worldwide, the EU AI Act introduces a risk-based regulatory framework for AI governance. Namely, AI systems are classified and covered by the Act depending on the level of risk they pose. In this respect, while certain use cases for AI systems are directly prohibited, most of the requirements and obligations are stipulated for high-risk AI systems (“HRAIS”). On the other hand, non-prohibited and non-high-risk AI systems are subject to either limited transparency rules or no obligations at all.

Under this risk-oriented legislative governance scheme, conformity assessments are one of the backbones of achieving a safe and trustworthy AI market. As mandated by the Act, all HRAIS must undergo a conformity assessment procedure before being marketed in the EU. This assessment is vital for ensuring the compliant development and deployment of AI systems. In this blogpost, we provide an overview of these conformity assessments under the EU AI Act.