Evidence reviewed as of 2026-07-13.
What happened
The European Commission has issued draft guidelines intended to assist providers and deployers of AI systems in understanding the classification of high-risk AI systems. These guidelines also aim to support market surveillance authorities in their roles.The primary goal of these guidelines is to facilitate a consistent application and effective enforcement of Article 6 of the AI Act.The document presents the Commission's interpretation of specific concepts related to classification.
What the primary source establishes
The draft guidelines provide the Commission's interpretation of key classification concepts and offer examples of AI systems that may or may not be considered high-risk. It is important to note that these examples are not exhaustive and may be updated over time. The table below summarizes key aspects of the draft guidance.
| Label | Finding | Basis Claim IDs |
|---|---|---|
| Purpose of Guidelines | To support providers, deployers, and market surveillance authorities in applying Article 6 of the AI Act. | C1, C2, C3 |
| Nature of Guidelines | They set out the Commission’s interpretation of classification concepts and include non-exhaustive examples. | C4, C5, C6 |
| High-Risk Scenarios | An AI system is considered high-risk in two scenarios: when it is a product safety component or when it falls under specific use cases listed in Annex III of the AI Act. | C7, C8 |
What it means for legal, compliance, and AI governance teams
For General Counsel and Compliance Leaders (CCO): AIGI analysis indicates these draft guidelines clarify obligations for providers and deployers regarding high-risk AI system classification under the EU AI Act. The guidelines aim to support both providers and deployers of AI systems, as well as competent market surveillance authorities, facilitating uniform application and effective enforcement of Article 6 of the AI Act. Compliance officers should note the non-exhaustive nature of examples, implying a need for continuous monitoring of updates and comprehensive internal assessments beyond the provided list. The document is currently in a consultation phase. What internal control mechanisms are in place to ensure ongoing alignment with evolving high-risk AI classification guidance, especially given the non-exhaustive examples?
For Chief AI Officers (CAIO): AIGI analysis suggests these draft guidelines are crucial for enterprise AI governance, directly impacting the risk classification of AI systems within a portfolio. An AI system is considered high-risk in two scenarios, including if it falls into use cases listed under Annex III of the AI Act. The clarification of high-risk scenarios, particularly those linked to Annex III use cases, requires Chief AI Officers to review existing and planned AI deployments for potential reclassification. The policy direction is tightening, requiring proactive planning. How will the organization's AI system portfolio be audited against the clarified high-risk classification criteria, especially concerning Annex III use cases, to ensure compliance?
For Chief Technology Officers (CTO): AIGI analysis indicates that these draft guidelines, while not yet final, provide critical insight into future technical compliance requirements for high-risk AI systems. The guidelines set out the Commission’s interpretation of certain classification concepts and contain examples of AI systems that should or should not be classified as high-risk. CTOs should consider the implications for system design and development, particularly regarding the need for robust classification mechanisms and the potential for updated examples, as these examples are not exhaustive and may be updated. This may influence build-vs-vendor decisions. What technical processes will be implemented to continuously assess and adapt AI system classifications in response to evolving guidance and non-exhaustive examples?
Evidence limits
The excerpt does not detail specific controls or practices required for compliance.
What to watch
As these are draft guidelines, their content is subject to change. The examples provided for high-risk classification are explicitly stated as non-exhaustive and may be updated over time. This suggests that organizations will need to maintain ongoing vigilance and adapt their internal classification processes as the guidance evolves.
Frequently asked questions
What is the purpose of the EU Commission's draft guidelines on high-risk AI classification?
The guidelines aim to support providers and deployers of AI systems, as well as market surveillance authorities, in uniformly applying and effectively enforcing Article 6 of the AI Act.
How does the draft guidance define a high-risk AI system?
An AI system is considered high-risk in two scenarios: if it is a safety component of a product or if it falls into specific use cases listed under Annex III of the AI Act.
Are the examples of high-risk AI systems in the guidelines exhaustive?
No, the examples of AI systems that should or should not be classified as high-risk are not exhaustive and may be updated over time.