The European Union has adopted the world’s first comprehensive artificial intelligence regulatory structure – the AI Act.
The Basics: The AI Act functions generally by organizing AI technology into one of four risk categories: (1) Prohibited AI; (2) High-Risk AI; (3) Limited Risk AI; and (4) Minimal Risk AI. Different requirements are then imposed on the AI based on the designated risk category.
Prohibited AI is just that – prohibited. Examples of prohibited AI include AI-deemed behavioral manipulation, predictive policing, untargeted scraping of facial images, and emotion recognitions systems. Prohibited AI violations face penalties of the lower of 7% of global turnover (revenue) or €35M.
HR AI Considered High-Risk AI: High-Risk AI, which includes Recruitment, HR, and Worker Management AI, carries several significant requirements, such as a fundamental rights and conformity assessment, data governance, and transparency. Violators face fines up to 3% of global turnover (revenue) or €15M, whichever is lower.
Not without detractors: The adoption of the AI Act drew immediate criticisms from French President Emmanuel Macron who warned the legislation risks hampering European technological companies and development compared to China and the United States.
What’s next: As HR Policy Global’s Tom Hayes writes, the method of adoption of the AI Act means that it does not need to be transposed into national laws. The AI Act still needs approval from the EU Council of Ministers and the Parliament – which is expected to happen in 2024. The subsequent compliance grace period will span from 6-24 months.
