Short answer: the EU AI Act (Regulation (EU) 2024/1689) is the European Union’s law on artificial intelligence. It sorts AI by risk, bans a few practices outright, sets strict rules for high-risk uses, and requires transparency for chatbots and AI-generated content. For most business websites, the part that matters is Article 50, which has applied since 2 August 2026: tell visitors when they talk to an AI, label deepfakes, and label unreviewed AI text on public-interest topics.
This guide explains the law in plain terms, the timeline after the 2026 Digital Omnibus changes, and what it means for a typical website.
What the EU AI Act is
The AI Act was adopted in 2024 and entered into force on 1 August 2024. It applies in stages. Unlike GDPR, which protects personal data, the AI Act regulates AI systems themselves: how they’re built, sold and used. It applies to providers (who build AI or put it into service under their name) and deployers (who use AI professionally), inside the EU and outside it when the AI’s output is used in the EU.
In 2026 the EU amended the Act through the Digital Omnibus on AI (Regulation (EU) 2026/1744), mainly to delay the high-risk rules and simplify some obligations. The transparency rules for websites were not delayed.
The four risk levels
| Level | What it means | Examples |
|---|---|---|
| Unacceptable risk | Banned | Social scoring, manipulative AI causing significant harm, untargeted scraping of face images, emotion recognition at work or school, AI apps that create non-consensual intimate images |
| High risk | Strict requirements: risk management, data quality, documentation, human oversight, registration | AI for hiring, credit scoring, education, essential services, safety components of products |
| Transparency risk | Disclosure duties (Article 50) | Chatbots, deepfakes, AI-generated content |
| Minimal risk | No specific AI Act duties beyond AI literacy | Spam filters, most recommendation and search features |
General-purpose AI models, such as the models behind ChatGPT, Claude or Gemini, have their own set of rules for the model makers.
Timeline
- 1 August 2024: entry into force.
- 2 February 2025: bans on prohibited practices and the AI literacy duty (Article 4) apply.
- 2 August 2025: rules for general-purpose AI models, governance and penalties apply.
- 2 August 2026: transparency obligations (Article 50) apply, along with most of the rest of the Act.
- 2 December 2026: end of the grace period for machine-readable marking by providers of generative AI already on the market, and for the new ban on AI “nudifier” apps.
- 2 December 2027: high-risk rules for stand-alone systems in Annex III areas such as employment and credit.
- 2 August 2028: high-risk rules for AI built into regulated products.
Why it matters for your website
Most websites aren’t building high-risk AI. But almost every business site now uses AI somewhere: a chatbot, AI-written blog posts, AI images, an AI writing or SEO plugin, recommendations in a shop. That makes the website owner a deployer, and Article 50 applies to several of those uses.
What you must do
- Chatbots: visitors must know they’re talking to an AI from the first interaction. See chatbot compliance.
- Deepfakes: realistic AI images, audio or video of people, places, objects or events that could pass as real need a visible label.
- AI text on public-interest topics: must be labelled unless a qualified person reviewed it and someone takes editorial responsibility. See do you need to disclose AI-generated content.
- Clear notices: disclosures must be clear, given at first exposure and accessible. See where to place AI disclosures.
- AI literacy: people who use AI on your behalf should understand its limits.
What you usually don’t need to do
- Label every AI-assisted text. Product descriptions and marketing copy typically fall outside the public-interest rule.
- Label stylised illustrations that nobody would take for a photo.
- Register anything or carry out a conformity assessment, unless you build high-risk AI.
Penalties
Fines go up to 35 million euros or 7% of worldwide turnover for banned practices, and 15 million euros or 3% for most other breaches, including Article 50. For SMEs, the lower amount applies. National authorities enforce the rules; in Germany that’s mainly the Bundesnetzagentur. Read EU AI Act fines and penalties explained for how amounts are set.
How it relates to GDPR
The two laws apply side by side. If your AI tool processes personal data, as most chatbots do, you need to meet both. Our EU AI Act vs GDPR comparison explains where they overlap.
Where to start
- Make a list of every AI feature and every kind of AI content on your site. Our website AI audit shows how.
- Check each item against the transparency rules for websites.
- Fix the gaps and keep a dated record.
On WordPress, the free EU AI Act Ready plugin handles the visitor-facing part: disclosure levels per post, notices only where needed, AI image labels, chatbot notices and an exportable compliance record. It runs on your own server and needs no account.
Frequently asked questions
Does the EU AI Act apply to small businesses?
Yes. There’s no size exemption for the transparency rules. SMEs get lower fine caps, simplified documentation for some obligations, and support measures.
Does it apply to websites outside the EU?
Yes, when the AI system’s output is used in the EU, for example a chatbot serving EU visitors.
Is using ChatGPT for my blog allowed?
Yes. The question is only whether a post needs a disclosure, which depends on its topic and whether a qualified person reviewed it.
Was the AI Act postponed?
Only partly. The Digital Omnibus delayed high-risk obligations to December 2027 and August 2028. Transparency obligations for chatbots, deepfakes and AI text have applied since 2 August 2026.