EU's New AI Transparency Rules Take Effect

The European Union has implemented new transparency regulations for artificial intelligence systems, effective August 2, 2026. These rules mandate that AI systems must inform users when they are interacting with a machine or viewing content that has been altered.

Furthermore, the European Commission's Office of Artificial Intelligence is now empowered to oversee generative AI models and levy fines against companies that fail to adhere to these obligations. This includes AI used in public services, customer interactions, and fraud reporting systems. Companies like OpenAI, Google, and Anthropic must design their systems to clearly indicate machine interaction.

Developers of AI systems used for emotion recognition or biometric categorization must also notify individuals when they are being analyzed, with exceptions for tools used in crime prevention and prosecution. Starting December 2, AI systems will be required to embed watermarks in manipulated text, photos, videos, and audio, enabling easier identification of machine-generated content by the public.

However, companies like ChatGPT and Google's Gemini, along with social media platforms with AI tools, will not be legally responsible for user-generated content; users themselves must disclose any manipulation. Fictional, artistic, or satirical content is exempt, but AI-generated content impacting public interest, such as election information, judicial processes, or public health matters, must be disclosed unless previously human-supervised.

Beyond transparency, the EU will prohibit the creation of non-consensual AI-generated sexualized images by December 2. This ban follows controversy surrounding AI-generated images. Platforms like X (formerly Twitter) will have until the end of the year to implement adequate protective measures.