# EU AI Act Enters Enforcement Phase New rules mandate AI content labeling by tech companies - **Source:** [INCRYPTED](https://brifing.si/source/7cd4625d-3f08-4cc1-983d-4c7d991839ba) - **Published:** 2026-08-29 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/1d40bc3e-ed91-4737-a048-1a1ab27dfdfa ## EU AI Act now in effect The European Union's Artificial Intelligence Act (AI Act) entered its enforcement phase on August 2, 2026. The new regulations impose transparency and disclosure requirements for chatbots and AI-generated content. The EU AI Office now has authority to sanction non-compliant companies. It can request access to AI models and demand information sharing from developers. ## Tech giants adapt to new rules Anthropic announced future Claude models will embed invisible watermarks in generated text, with APIs for detection. Google and Meta are developing content labeling and transparency tools for AI-generated material. OpenAI publishes summaries of training data and implements content origin signals. Microsoft is adjusting its internal risk assessment processes to comply with AI Act requirements. ## New user notification system A key provision requires users to be notified when interacting with AI or encountering AI-generated content. Patrick van Ecke from Cooley law firm in Brussels describes the regulation as complex legislation. Users now receive warnings when viewing AI-produced content. This marks the first time ordinary users can directly experience the law's impact in their digital interactions. ## Watermarks as legal protection Amy Worley, Data Protection Director at Berkeley Research Group, notes AI usage notifications are becoming market standard. She suggests watermarks could serve as legal protection in disputes. Worley emphasizes the EU AI Act provides much-needed standards in this rapidly evolving field. The regulations are expected to have global ripple effects beyond the EU market. ## Phased implementation timeline The August 2026 rules are just the beginning. Regulations for high-risk AI systems in education, biometrics, and migration will take effect in December 2027. AI integrated into physical products faces requirements starting August 2028. Companies like Microsoft are already adapting their governance and risk assessment frameworks to meet these upcoming obligations. ## Potential global standard Experts predict the EU AI Act may follow GDPR's path to become a global benchmark. The EU's leadership in digital regulation positions its rules as potential de facto standards worldwide. This could extend the Act's influence beyond EU borders, affecting global tech companies' operations. The regulations may solidify Europe's role as a digital policy norm-setter. ## Corporate compliance costs Meeting the new requirements imposes significant costs on companies. Developing transparency mechanisms, content labeling systems, and risk assessment processes presents challenges. Smaller AI developers may face disproportionate burdens. While ensuring accountability, the regulations must balance innovation needs. The long-term impact on AI development remains to be seen. ## New era of digital regulation The EU AI Act represents a major step toward responsible AI development, prioritizing user rights and corporate accountability. Its success depends on practical implementation and industry adaptation. The regulations aim to prevent harmful AI uses while fostering innovation. As the first comprehensive AI law, it sets important precedents for future governance of emerging technologies. ## Sources covering this story - https://incrypted.com/zakon-es-ob-yy-pereshel-k-etapu-realnoho-kontrolya-novye-pravyla-uzhe-deystvuyut/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._