Executive summary – EU Regulation on AI
The EU-regulation on trustworthy AI was presented on April, 21 2021, based on the intentions of the EU’s Ethical guidelines for trustworthy AI.
The regulation takes a risk based approach targeting primarily AI systems that are deemed to have unacceptable risk and high-risk. It targets both developers and providers, as well as users of AI-systems. While it contains very light regulation for AI-systems with limited risk, transparency obligations only, it encourages voluntary Codes of Conduct for all AI-systems.
AI systems deemed to entail unacceptable high risk and therefore will be banned are AI systems considered a clear threat to the safety, livelihoods and rights of people, from social scoring by governments to toys using voice assistance that encourages dangerous behaviour.
High-risk AI systems include AI-technology used in, inter alia, critical infrastructure (e.g. transport) that could put the life and health of citizens at risk, law enforcement, educational training (that may determine the access to education by scoring of exams), essential public and private services (e.g. credit scoring when granting credit or providing public aid), employment (e.g. CV-sorting software for recruitment purposes), migration, asylum and border control. High-risk AI systems will be subject to strict obligations before they can be put on the market:
- Adequate risk assessment and mitigation systems
- High quality of the datasets feeding the system to minimise risks and discriminatory outcomes
- Logging of activity to ensure traceability of results
- Detailed documentation providing all information necessary on the system and its purpose for authorities to assess its compliance
- Clear and adequate information to the user
- Appropriate human oversight measures to minimise risk
- High level of robustness, security and accuracy
The regulation contains high fines for non-compliance of up to six percent of a company’s annual turnover.
The proposal will be reviewed by the European Parliament and the European Council with a likely entry-into-force date for the regulation in 2024.