On August 2, 2026, the European Union’s AI Act entered a new enforcement phase as Article 50—covering transparency obligations for AI-generated content—became fully applicable. This marks a pivotal moment in AI regulation, expanding legal accountability beyond model developers to include those deploying AI systems.
Under Article 50, AI providers must ensure that synthetic audio, image, video, or text outputs are marked in a machine-readable format, such as watermarks or metadata, to clearly indicate their artificial origin (en.wikipedia.org). Meanwhile, deployers—those who publish or use AI-generated content—are required to disclose when content is AI-generated, especially if it informs the public on matters of public interest, unless it has undergone human review with editorial responsibility (reddit.com).
The enforcement of Article 50 comes amid broader delays in the AI Act’s high-risk provisions. The Digital Omnibus regulation, which entered into force on July 27, 2026, postponed obligations for high-risk AI systems: standalone high-risk systems now face compliance deadlines of December 2, 2027, and embedded systems by August 2, 2028 (consilium.europa.eu). However, Article 50’s transparency rules were not delayed and are now binding.
This shift places immediate compliance pressure on a wide range of organizations. Deployers—regardless of whether they built the AI system—must now clearly label AI-generated content, including chatbots, deepfakes, and AI-authored text, or face potential penalties. The law’s reach extends to any entity publishing such content within the EU, echoing GDPR’s extraterritorial scope (reddit.com).
Enforcement mechanisms are now active. National market surveillance authorities have investigatory powers, and the EU AI Office can impose fines up to €15 million or 3% of global turnover for general-purpose model violations, and up to €35 million or 7% for prohibited practices (reddit.com). While some observers question how aggressively enforcement will be pursued—citing limited regulatory staffing and recent rule clarifications—others argue that early enforcement could set a strong precedent for compliance culture (reddit.com).
For businesses and developers, the takeaway is clear: transparency obligations are no longer optional. Immediate action is required to update user interfaces, content workflows, and documentation practices to ensure AI-generated outputs are properly labeled and traceable. As the EU tightens its regulatory grip, organizations must treat Article 50 compliance as a present-day imperative—not a future concern.