The EU AI Act for marketers: Article 50, labelling AI content and the key dates

What the AI Act is, and why Article 50 is the marketer's article

The Artificial Intelligence Act, Regulation (EU) 2024/1689, is the European Union's horizontal law on AI, adopted on 13 June 2024. Most of its weight falls on prohibited practices, high-risk systems and general-purpose AI models. For most marketing teams, though, the part that changes day-to-day work is Article 50: transparency obligations for providers and deployers of certain AI systems.

The European Commission has published non-binding guidelines on these transparency obligations (20 July 2026) and a voluntary Code of Practice on transparency of AI-generated content (10 June 2026). We draw on both below.

Important: this is a practical overview for marketers, not legal advice. Whether a specific campaign, tool or piece of content is in scope depends on facts, and you should check with qualified counsel.

Providers and deployers: which one are you?

Article 50 splits duties between providers (who develop an AI system or place it on the market under their name) and deployers (who use one in a professional capacity). The Commission's guidelines make several points relevant to marketing:

• Businesses using AI systems under their authority are deployers, unless the use is purely personal. Authority means deciding to use the system and how; it does not require technical control.
• A business that builds its own chatbot and runs it under its own name is a provider of that chatbot.
• A company that merely commissions an advertising agency, without controlling whether and how AI is used, is not the deployer for that work.
• Non-EU companies are covered where they foresee their output being used in the EU, which matters for Chinese brands advertising in Europe.

The four Article 50 obligations, in marketing terms

1. Chatbots and AI interactions (Article 50(1), providers). Systems that interact directly with people must be designed so people know they are dealing with AI, unless that is obvious to a reasonably well-informed person. The guidelines read the "obvious" exception narrowly: website and helpdesk chatbots whose replies could pass as human-written do not qualify. Disclosure should be prominent and in plain language at the first interaction; a line buried in terms and conditions, or a generic "this site uses AI", is not enough on its own.
2. Machine-readable marking (Article 50(2), providers). Providers of tools that generate synthetic audio, images, video or text must mark outputs in a machine-readable way so they can be detected. This mainly concerns the tool vendors you buy from. Purely assistive editing that does not substantially alter the input is excluded. Generative systems already on the market before 2 August 2026 have until 2 December 2026 to comply.
3. Emotion recognition and biometric categorisation (Article 50(3), deployers). If you use such systems, for example in retail analytics, people exposed to them must be informed, and data protection law applies.
4. Deepfakes and public-interest text (Article 50(4), deployers). You must disclose that image, audio or video deepfakes are artificially generated or manipulated. For evidently artistic, creative, satirical or fictional work, disclosure can be done in a way that does not spoil the work. You must also disclose AI-generated or manipulated text published to inform the public on matters of public interest, unless it has undergone human review or editorial control and someone holds editorial responsibility for publication.

Advertising is not exempt. The guidelines give examples such as an AI-generated video of a celebrity influencer in a promotional context, a realistic AI avatar of a company's CEO, and an AI-generated product image that misrepresents what the real product looks like. A real product placed against an AI-generated background is not treated as a deepfake if it does not mislead about the product itself. Article 50(5) requires all these disclosures to be "clear and distinguishable" and given at the latest at first interaction or exposure; deployers cannot rely on invisible machine-readable marks alone.

The timeline

• 1 August 2024: the AI Act entered into force.
• 2 February 2025: prohibitions and general provisions, including AI literacy, began to apply.
• 2 August 2025: obligations for general-purpose AI model providers and the governance framework applied.
• 2 August 2026: most rules, including Article 50 transparency, apply and enforcement starts.
• 2 December 2026: end of the Article 50(2) transitional period for generative systems already on the market; new prohibitions on AI generating non-consensual sexual deepfakes and child sexual abuse material apply.
• 2 December 2027 and 2 August 2028: high-risk rules for Annex III systems and for AI in regulated products respectively.

The December dates were introduced by the Digital Omnibus on AI, which, as Regulation (EU) 2026/1744, entered into force on 27 July 2026. The Commission's AI Act Service Desk timeline reflects these amended dates. According to law firm Cooley, content generated and published before 2 August 2026 does not need to be labelled retroactively.

Penalties are set in Article 99: breaches of Article 50 can lead to fines of up to €15 million or 3% of worldwide annual turnover, whichever is higher. For SMEs and start-ups, whichever is lower applies.

A compliance checklist for marketing teams

1. Inventory your AI use. List every tool that generates content, every chatbot, and any emotion or biometric analytics.
2. Classify your role per tool. Are you a provider (your own branded chatbot) or a deployer (using a vendor's tool)? Clarify roles with agencies in writing.
3. Fix chatbot disclosure. Show a plain-language notice at the start of every AI conversation and make sure the bot confirms it is AI when asked.
4. Set a deepfake rule. Any realistic depiction of real people, places, products or events created or altered with AI gets a visible or audible label. Do not use AI to change how a real product looks.
5. Decide your text policy. For articles, news-style posts and statements on public-interest topics, either label AI-generated text or ensure documented human editorial review with a named responsible person.
6. Check your vendors. Ask tool providers how they meet Article 50(2) marking and whether they follow the Code of Practice.
7. Document and train. Keep records of decisions, approvals and labels, and include Article 50 in your AI literacy training.

Limits and open questions

The guidelines and the Code of Practice are not binding law; courts and national authorities will shape how terms such as "matters of public interest" apply to commercial content. Article 50 also sits alongside consumer and data protection rules, which may require further disclosures.

FAQ

Sources

• Regulation (EU) 2024/1689 (Artificial Intelligence Act) — EUR-Lex, Publications Office of the EU, 2024
• Article 50: Transparency obligations for providers and deployers of certain AI systems — AI Act Service Desk, European Commission, consolidated 2026
• Article 99: Penalties — AI Act Service Desk, European Commission
• Timeline for the implementation of the EU AI Act — AI Act Service Desk, European Commission
• Guidelines on transparency obligations for providers and deployers of AI systems — European Commission, 2026
• Code of Practice on transparency of AI-generated content: FAQ — European Commission, 2026
• Digital Omnibus on AI enters into force — ERTICO – ITS Europe, 2026
• EU AI Act: Transparency Obligations Take Effect 2 August 2026 — Cooley, 2026