The "high-risk" category is at the heart of the AI Act. These systems, which can have a significant impact on people's fundamental rights, are subject to a very strict set of obligations. But how do you know if your systems are affected?
What is a high-risk AI system?
An AI system is considered "high-risk" if it meets one of the following two criteria:
Caution: even if your system is not explicitly listed, it may qualify as high-risk if it is used to make important decisions affecting people.
The 8 categories of Annex III
Annex III precisely defines the areas where AI systems are presumed high-risk:
- blog_article3_cat1_item1
- blog_article3_cat1_item2
- blog_article3_cat1_item3
- blog_article3_cat2_item1
- blog_article3_cat2_item2
- blog_article3_cat3_item1
- blog_article3_cat3_item2
- blog_article3_cat3_item3
- blog_article3_cat4_item1
- blog_article3_cat4_item2
- blog_article3_cat4_item3
- blog_article3_cat5_item1
- blog_article3_cat5_item2
- blog_article3_cat6_item1
- blog_article3_cat6_item2
Obligations for high-risk systems
If your system qualifies as high-risk, you must comply with the following requirements:
Conformity assessment
Before placing a high-risk system on the market, you must carry out a conformity assessment. Depending on the case, this assessment may be:
blog_article3_alert_text
Is your system high-risk?
Our automated analysis tool evaluates your system and identifies its classification.
Start free auditThe "high-risk" qualification entails significant but not insurmountable obligations. The key is to anticipate and implement appropriate processes now.
