What are high-risk artificial intelligence systems
Among the various categories introduced by EU AI regulation, high-risk systems are likely the most relevant for businesses. Unlike prohibited applications, these systems can be developed and used, but must comply with specific requirements.
Introduction
Among the various categories introduced by the European regulation on artificial intelligence, that of high-risk systems is probably the most relevant for companies.
Unlike prohibited applications, these systems can be developed and used, but must comply with a series of specific requirements designed to reduce the risks associated with their use.
The definition and regulation of these systems are one of the central elements of the Artificial Intelligence Act.
When an AI system is considered high-risk
An artificial intelligence system is classified as high-risk when its use can have a significant impact on people's safety or their fundamental rights.
This can happen, for example, when AI is used to support or automate decisions that affect economic opportunities, access to services, or working conditions.
In these cases it is necessary to ensure that the system is developed and used in a reliable and controllable manner.
Why this category is so important
Many of the most widespread applications of artificial intelligence in organizations can fall into this category.
Think, for example, of systems used for candidate selection, for credit assessment, or for data analysis in the healthcare sector.
In these contexts, an error or algorithmic distortion can have significant consequences for the people involved.
Conclusion
For this reason, European regulation introduces specific requirements for high-risk AI systems.
Understanding whether a particular application falls into this category represents one of the most important steps for companies that use artificial intelligence.