EU AI Act

The EU AI Act is Regulation (EU) 2024/1689, the first broad horizontal law governing artificial intelligence. It entered into force on 1 August 2024 and applies in phases. Most online shops are not deploying “high-risk” AI, so the heavy compliance machinery does not apply to them. The part that does matter to merchants is the transparency obligations in Article 50.

The phased timeline

• 1 August 2024 — the Regulation enters into force.
• 2 February 2025 — prohibited AI practices become applicable (a short list of banned uses such as certain manipulative or social-scoring systems).
• 2 August 2025 — obligations for general-purpose AI models begin to apply.
• 2026–2027 — most high-risk system rules become applicable.

The phasing matters because it means the obligations are not a single switch. But for a typical e-commerce merchant the practical question is narrower: what do the transparency rules require of a shop that uses a support chatbot or AI-generated product content?

Article 50: transparency duties for merchants

Article 50 sets out transparency duties that can apply to ordinary businesses, not just AI developers. Three of them are relevant to online shops.

1. Tell people when they are interacting with an AI system

If your shop runs a support chatbot or AI-driven assistant, users must be informed that they are interacting with an AI system — unless that is already obvious from the circumstances. A short, clear label (“You’re chatting with our AI assistant”) satisfies the duty and avoids misleading customers.

2. Disclose AI-generated or manipulated media and public-information text

AI-generated or manipulated image, audio or video content — so-called “deep fakes” — must be disclosed as artificially generated or manipulated. The same applies to AI-generated text published to inform the public on matters of public interest. For a shop, this most often bites where you use AI to create product imagery or synthetic media that a reasonable customer would take to be a genuine photograph.

3. Machine-readable marking of synthetic content

Providers of AI systems that generate synthetic audio, image, video or text must mark the output in a machine-readable way so it can be detected as artificially generated. If you generate content in-house you should preserve any such markings; if you use a third-party tool, choose one that applies them.

What this means in practice

• Label your chatbot clearly as AI.
• Disclose AI-generated product images or media where they could be mistaken for real photographs.
• Keep or apply machine-readable markers on synthetic content.
• Remember these transparency duties sit alongside existing consumer-protection rules against misleading practices.

Most shops will not touch the high-risk regime at all. Focus on the transparency basics and revisit as later phases apply. See our FAQ on AI transparency for online stores, and the wider consumer-law modernisation context.