Subjects
Annual Updates Subcategories
Anti-Bribery & Compliance
Anti-Trust & Competition
Data Protection & Privacy Subcategories
Disputes and Grievances
Employment Law
Employment Status Subcategories
Equality, Diversity and Equal Rights Subcategories
Industrial & Workplace Relations Subcategories
International Mobility Subcategories
Pensions & Benefits
Protection of Business Interests Subcategories
Recruitment, Termination of Employment Subcategories
Remuneration Subcategories
Social Security Contributions
Terminations Subcategories
Working Hours & Pay Subcategories
Workplace Culture Subcategories
Jurisdictions
Americas Subcategories
APAC Subcategories
Cross Border
EMEA Subcategories
Global

EU: The EU AI Act – New Regulation

By - - AI Compliance

The EU Regulation on Artificial Intelligence (“The EU AI“) comes into effect incrementally between now and August 2nd 2026.

What has come in effect as off February 2nd 2025?

The EU AI lists “prohibited practices” that it regards as creating unacceptable risks. These had to be removed from the market and/or discontinued completely, by February 2nd 2025.

This restriction applies to all operators, whether as developer, provider or deployer.  It covers AI systems that:

  • Use social scoring techniques to classify a person or group based on behaviour or personal characteristics;
  • Use subliminal, manipulative or deceptive techniques that might materially distort the behaviour of a person or group;
  • Exploit vulnerability characteristics of people or groups to distort behaviour;
  • Create or expand facial recognition databases using images from the internet or CCTV;
  • Infer emotions of persons in a workplace or place of education;
  • Use profiling techniques or assessment of personal traits to predict the risk of criminal behaviour;
  • Use biometric systems that categorise individuals based on biometric data to deduce or infer race, opinions, trade union membership, religious beliefs or sexual orientation.
  • Use biometric real-time identification systems in public places. for law enforcement.                
Obligation on Qualifying Employers

Employers who count as providers (i.e. developers) or deployers (i.e. users) must take care to ensure  their employees and anyone working on AI systems on their behalf, have an adequate level of “AI literacy”.

EU guidelines on the meaning of AI literacy are expected shortly. It is understood to mean that staff have a sufficient level of education in the comprehension, application, oversight and critical evaluation of specific AI applications. This will include technical expertise, experience, education and training and an demonstrable understanding of the impact of AI usage.

Penalties

These will be applied from August 2nd, 2025 and depend on the nature and extent of the violation, as well as the business sector. The possible penalties are fines are up to 7% of global turnover, or fixed between EUR 7.5 million and EUR 35 million.

This is a high level general update only. Legal advice should be obtained on specific circumstances.

Subject: AI Compliance
Scroll to Top