The European Union’s artificial intelligence regulations enter their most consequential enforcement phase today, with the European Commission’s AI Office and national authorities beginning to enforce Article 50 of the AI Act—a transparency mandate affecting millions of AI applications across the bloc and beyond.
Starting August 2, 2026, the new transparency rules require AI systems to disclose their nature to users. Chatbots and interactive AI systems must inform users they are communicating with AI, not a human. Deepfakes—images, videos, or audio edited or generated using AI—must be clearly labeled. AI-generated or manipulated content must also carry machine-readable marks so it can be detected more easily.
The rules apply to both providers—companies that develop and release AI systems—and deployers, the organizations that use those systems in their operations. A media company publishing AI-generated news articles, a bank deploying a third-party chatbot on its customer-service portal, or a recruitment firm using AI-powered resume screening all fall under the requirements. The transparency obligations apply instantly to all AI systems in scope, regardless of when they were placed on the market.
The European Commission published guidelines on July 20 clarifying what compliance looks like across four key areas: direct interaction with users, AI-generated synthetic content, emotion recognition and biometric categorization systems, and deepfakes or AI-generated text on matters of public interest. The Commission also released a list of more than 180 organizations that have signed the Code of Practice on Transparency of AI-Generated Content, a voluntary framework that operationalizes the rules and provides practical technical measures for marking and labeling.
Non-compliance carries steep penalties. Organizations that fail to meet the transparency obligations face fines of up to €15 million or 3 percent of worldwide annual turnover, whichever is higher. Unlike most other AI Act provisions, these obligations are instantly applicable and do not include extended grace periods. Content generated before August 2 but published afterward must still comply with labeling requirements.
The August 2 deadline marks the culmination of a phased rollout that began in February 2025 with bans on prohibited AI practices. The AI Act, which entered into force on August 1, 2024, implements its provisions on a rolling basis through 2028, but the transparency rules represent the first broad-based enforcement action targeting the vast majority of organizations deploying generative AI. According to legal experts, the transparency tier is by far the most widely applicable section of the Act, extending beyond just AI developers to every business using these systems in a professional capacity.
The enforcement action reflects the EU’s stated goal of reducing deception, misinformation, and manipulation while safeguarding democratic processes and public trust. As generative AI tools have proliferated since late 2022, calls for clear labeling have grown louder. The original AI Act proposal included transparency obligations, which were expanded in response to the emergence of large-scale systems like ChatGPT and have largely remained in the final text adopted in June 2024.
Businesses that have not yet audited their AI systems or updated their compliance frameworks face immediate risk. The transparency obligations do not exempt systems released under free and open-source licenses, nor do they apply only to new deployments. Any organization using AI for professional purposes—whether a startup or a multinational—must ensure its systems comply by today. The Commission has indicated that signatories of the voluntary Code of Practice will benefit from a presumption of conformity, offering one practical pathway to demonstrate alignment with the rules.
Sources
- European Commission Digital Strategy — official press release on enforcement beginning August 2, 2026, and transparency requirements for chatbots, deepfakes, and AI-generated content
- Stibbe LLP — legal analysis of Article 50 transparency obligations, scope of providers and deployers, four categories of obligations, and enforcement penalties of €15 million or 3% of global turnover
- AI-Act.eu — implementation timeline and confirmation that transparency rules apply from August 2, 2026
- AI Act Service Desk (European Commission) — details on deployer obligations for deepfake disclosure and AI-generated text labeling
