On August 2, 2026, any AI system that interacts with humans, generates synthetic content, performs emotion recognition, or creates deepfakes must comply with EU AI Act Article 50 transparency obligations. Non-compliance can result in fines up to EUR 15 million or 3% of global turnover.
EU AI Act: 46 Days Until Transparency Rules Take Effect
What happens on August 2, 2026
The EU AI Act (Regulation 2024/1689) entered into force on August 1, 2024. Its obligations are being phased in over four years. The next major deadline is August 2, 2026, when transparency obligations under Article 50 become legally enforceable across all 27 EU Member States.
This is not a high-risk AI deadline. The Annex III high-risk requirements were postponed to December 2, 2027 by the Digital Omnibus Agreement of May 7, 2026. August 2 is specifically about transparency.
What Article 50 requires
Article 50 establishes four categories of transparency obligations:
- AI-to-human interaction: Any AI system designed to interact directly with people must disclose that the user is communicating with an AI. The notification must be clear and timely — not buried in terms of service.
- Synthetic content: AI-generated text, audio, images, and video must be marked as artificially generated using machine-readable formats. This includes watermarking and metadata standards such as C2PA.
- Emotion recognition and biometric categorisation: Deployers of such systems must inform the individuals being assessed, before the system is used.
- Deepfakes: Any person using AI to generate or manipulate content that resembles existing persons, places, or events must disclose that the content has been artificially generated or manipulated.
What the penalties look like
Failing to comply with Article 50 transparency obligations falls under the EU AI Act's Tier 2 penalty structure:
| Penalty tier | Maximum fine | Applies to |
|---|---|---|
| Tier 1 (Prohibited practices) | EUR 35,000,000 or 7% of global turnover | Article 5 violations |
| Tier 2 (Other obligations) | EUR 15,000,000 or 3% of global turnover | Article 50 transparency, GPAI, high-risk requirements |
For SMEs and startups, each fine is capped at the lower of the percentage or the fixed amount, not the higher — providing relative protection for smaller operators.
What to do now
If you have not started, begin with an AI system inventory. Map every AI system your organisation develops, deploys, or procures against the four Article 50 categories.
For each system with transparency obligations:
- Implement human-interaction disclosure (chatbots, virtual assistants, AI phone systems)
- Add machine-readable metadata to AI-generated content (C2PA, watermarking)
- Establish notification procedures for emotion recognition or biometric systems
- Create deepfake disclosure workflows for synthetic media
Document every compliance measure. Under the AI Act, if it is not documented, it cannot be demonstrated to a regulator.
Are you ready for August 2? Check now.
Free AI Act Readiness Check 3 minutes — 10 questions — instant resultsThis article is for informational purposes only and does not constitute legal advice. Regulatory requirements change frequently — verify current rules with official sources. Published June 17, 2026 by Sawai Gyoseishoshi Office, Hiroshima, Japan.