Europe's landmark rules on artificial intelligence (AI) will come into force next month after EU countries ratified a political deal reached in December, which will limit the potential of the technology used in business and daily life. This will set a global benchmark.
The European Union's AI law is more comprehensive than the United States' lighter voluntary compliance approach, while China's approach aims to maintain social stability and state control.
The vote by EU countries came two months after MEPs endorsed the AI Bill drafted by the European Commission in 2021, with a number of important changes.
Concerns about AI contributing to misinformation, fake news, and copyrighted content have grown globally in recent months. These concerns come as generative AI systems such as Microsoft-backed OpenAI's ChatGPT and Google's chatbot Gemini grow in popularity.
“This groundbreaking law, the first of its kind in the world, addresses global technological challenges and also opens opportunities for societies and economies,” Belgian Digitalization Minister Mathieu Michel said in a statement. ” he said.
“With the AI Act, Europe will emphasize the importance of trust, transparency and accountability when dealing with new technologies, while ensuring that this rapidly changing technology can thrive and foster innovation in Europe,” he said. said.
While the AI Act imposes strict transparency obligations on high-risk AI systems, it relaxes such requirements for general-purpose AI models.
This would limit the government's use of real-time biometric surveillance in public places to cases of certain crimes, preventing terrorist attacks, and searching for suspects in the most serious crimes.
Patrick van Ecke from law firm Cooley said the new law would have an impact beyond EU member states.
“This law will have a global impact. Non-EU companies that use EU customer data in their AI platforms will need to comply. Other countries and regions will also “We are likely to use the AI Act as a blueprint,” he said, referring to EU privacy rules.
The new law will apply from 2026, but the ban on the use of artificial intelligence in social scoring, predictive policing and untargeted scraping of facial images from the internet and CCTV footage will begin within six months of the new regulations coming into force.
Mandates for general purpose AI models will apply after 12 months, and rules for AI systems embedded in regulated products will apply after 36 months. Fines for violations range from 7.5 million euros (1.5% of global turnover) to 35 million euros (7% of worldwide turnover), depending on the type of violation.
