5 things you need to know about the new AI regulation that could reshape the entire internet
- The European Union has officially passed the final version of its landmark AI Act, the world's first comprehensive legal framework governing artificial intelligence. This means that any company using or developing AI for the global market must comply with strict rules on transparency, risk management, and human oversight, or face fines of up to 7% of global annual revenue.
- The regulation categorizes AI systems into four risk levels: unacceptable, high, limited, and minimal. Unacceptable risk systems, like social scoring by governments and real-time biometric surveillance in public, are banned outright. High-risk systems, such as those used in hiring, credit scoring, and critical infrastructure, must undergo a rigorous conformity assessment before release.
- A major surprise in the final draft is the requirement for all generative AI models, including chatbots and image generators, to clearly label their output as AI-generated. This specifically targets deepfakes and synthetic media, and it forces platforms like TikTok, Instagram, and X to add watermarks and metadata to any AI-created content.
- The enforcement timeline is staggered to give companies time to adapt. The bans on unacceptable risk take effect in six months, while rules for high-risk systems have up to 24 months to be implemented. However, open-source AI models and research projects are largely exempt, which has created a heated debate about fairness between big tech and smaller developers.
- Global tech giants like OpenAI, Google, and Meta are already scrambling to update their policies and designate compliance officers, fearing disruption to their European user base of over 450 million people. The ripple effect is expected to push countries like Japan, Canada, and Brazil to adopt similar laws, making this a potential global standard for artificial intelligence news.