EU’s AI Act sets strict guidelines for companies, enforcement starts Feb 2025

The European Union’s Artificial Intelligence (AI) Act, which was officially adopted in August 2024, is set to bring sweeping changes to how businesses handle AI technologies. As of February 2, 2025, companies operating within the EU will need to comply with specific regulations to ensure AI literacy and eliminate certain prohibited AI practices, or face substantial fines.

The AI Act introduces a legal framework that applies to the development, deployment, and use of AI systems in the EU. Aimed at safeguarding public trust and promoting responsible AI, the regulation is divided into several chapters, with Chapters I and II taking effect in early 2025.

AI Literacy Requirements (Chapter I)
One of the core aspects of Chapter I is the mandate for AI literacy. Companies providing or deploying AI systems will be required to ensure their staff and anyone interacting with AI technologies understand its risks, opportunities, and potential harms. To comply, businesses must organize regular training sessions to boost AI literacy across their workforce. The goal is to ensure that everyone involved in the use and operation of AI is adequately equipped to handle these technologies safely and responsibly.

Prohibited AI Practices (Chapter II)
Chapter II outlines a range of AI practices that are considered harmful and will be prohibited once the regulation becomes enforceable. These practices include AI systems that use subliminal or manipulative techniques to distort behavior, as well as systems that exploit vulnerabilities in individuals or specific groups for harmful purposes. Additionally, AI systems that track or infer emotions in sensitive settings such as workplaces or educational institutions are banned. Another significant restriction involves AI technologies that create or expand facial recognition databases without proper consent, such as scraping images from the internet or using public CCTV footage.

Fines for Non-Compliance
Companies failing to adhere to these regulations face steep penalties. Violations of prohibited AI practices could result in fines up to €35 million or 7% of a company’s global annual turnover, whichever is higher. Non-compliance with AI literacy requirements could also lead to sanctions, although the exact penalties will be determined by individual EU member states.

As the February 2025 deadline approaches, businesses are urged to begin preparing their teams for the upcoming changes, ensuring compliance with both the literacy and prohibited practices requirements to avoid heavy fines.

Also Read: Google challenges EU's $4.5 billion antitrust fine over android practices