The EU AI Act sets out rules for how AI systems may be developed and used in the EU. The law is risk-based: certain applications are prohibited, high-risk systems face stricter requirements, and others are permitted but must meet transparency obligations in specific situations.
On August 2, 2026, Article 50 takes effect. It requires that people be informed when they interact with AI or encounter AI-generated content. Article 50 sets different obligations depending on whether your company is a provider or a deployer.
Provider: a natural or legal person, public authority, agency or other body that develops an AI system or a general purpose AI model (or that has an AI system or a general purpose AI model developed) and places them on the market or puts the system into service under its own name or trademark, whether for payment or free of charge.
Deployer: any natural or legal person, public authority, agency or other body using an AI system under its authority except where the AI system is used in the course of a personal non-professional activity.
For most mid-sized companies using AI within their own business operations, the AI Act considers them deployers. Companies that build and offer AI systems to others are providers.
What is required?
Article 50 establishes transparency requirements in four main cases.
1. When people interact directly with AI
When an AI system interacts directly with people (chatbots, phone assistants, virtual assistants, automated customer support), those people must be told they are interacting with AI. The notice must be given at the start of the interaction, clearly and prominently. It should not be buried in terms and conditions or a footer.
2. When AI generates synthetic content
Providers of generative AI systems must make AI-generated or AI-manipulated outputs identifiable in a machine-readable format, across text, images, audio, and video. This applies to providers offering generative AI systems to others, not to companies using such systems internally.
The Code of Practice on Transparency of AI-Generated Content, published June 10, 2026, provides voluntary guidance on implementation, including EU icons for labelling AI-generated content.
3. When emotion recognition or biometric categorisation is used
If your company uses AI to analyse emotions, mood, stress, or demographic characteristics, affected people must be clearly informed. Assess the use case carefully before deployment. Some uses are already prohibited under Article 5 of the AI Act, and these systems may also trigger data protection obligations.
4. When AI creates deepfakes or certain texts of public interest
AI-generated or manipulated images, audio, or video that could appear authentic must be disclosed as AI-generated. Text published to inform the public on matters of public interest may also require disclosure.
The obligation does not apply where the AI-generated content has undergone a process of human review or editorial control and a natural or legal person holds editorial responsibility for the publication.
Key dates
| Date | What applies |
|---|---|
| June 2026 | Code of Practice on AI-Generated Content published |
| August 2, 2026 | Article 50 transparency obligations apply to providers and deployers |
What Companies Should Do Now
All organisations
- Map the AI systems in use and identify which may fall under Article 50.
- Use the EU AI Act Compliance Checker to assess which obligations apply to your specific systems.
- Review the Code of Practice for practical implementation guidance.
- Require AI providers to guarantee compliance contractually.
Providers
- Ensure AI-interactive systems display a clear notice at the point of first interaction.
- Implement machine-readable labelling of synthetic outputs across text, images, audio, and video.
Deployers
- For emotion recognition or biometric categorisation, check against Article 5 prohibitions before considering disclosure under Article 50.
- Map any use cases that produce realistic synthetic media and design appropriate disclosure.
- For published AI-generated text on matters of public interest, either disclose or establish a human review process with editorial responsibility.
What this means in practice
The practical task before August 2 is to identify where AI is in use, whether your company acts as a provider or deployer, and whether a transparency or disclosure obligation follows.
The highest-priority areas are customer-facing AI systems, phone assistants, chatbots, synthetic media, AI-generated public content, and emotion or biometric analysis.
If you want to assess where AI can be deployed safely in your business, contact brainbot. We help mid-sized companies design and implement AI agents with clear workflows, permissions, and data control.
Sources
EU Artificial Intelligence Act. “The EU AI Act’s Transparency Rules: A Practical Guide to Article 50.” https://artificialintelligenceact.eu/transparency-rules-article-50/
European Commission. “Code of Practice on Transparency of AI-Generated Content.” https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content
European Commission. “EU Icons for Labelling AI-Generated Content.” https://digital-strategy.ec.europa.eu/en/policies/eu-icons-labelling-ai-generated-content
EU AI Act. “EU AI Act Compliance Checker.” https://artificialintelligenceact.eu/assessment/eu-ai-act-compliance-checker/