The EU Commission has published an overview of “AI practices that are deemed unacceptable due to their potential risks to European values and fundamental rights” (here).

The guidelines provide legal explanations and practical examples to help stakeholders understand and comply with the AI Act's requirements for prohibited systems under Article 5.   They reflect the Commission's current interpretation of the prohibitions, but will not be binding until formally adopted by the Commission at a later date. It is only from that moment that the guidelines will be applicable.  Further, application will require a case-by-case assessment, which takes due account of the specific situation at issue in an individual case. Therefore, the examples given in the guidelines are merely indicative.

Prohibited AI systems include:

1. Harmful manipulation, and deception
2. Harmful exploitation of vulnerabilities
3. Social scoring
4. Individual criminal offence risk assessment and prediction
5. Untargeted scraping to develop facial recognition databases
6. Emotion recognition
7. Biometric categorisation
8. Real-time remote biometric identification.

The guidelines should be read in conjunction the EU's guidelines on what constitutes an AI system (under Article 3).  Only AI systems meeting the definition will be subject to the AI Act, including the prohibitions. Read our summary here.

To help understand whether the EU AI Act applies, check out our flowchart to help navigate key aspects of the Act (here).

If you would like to discuss how current or future regulations impact what you do with AI, please contact Tom Whittaker, Brian Wong, Lucy Pegler, Martin Cook, Liz Smith or any other member in our Technology team.