EU AI Act: what changes for your chatbot on August 2, 2026

On August 2, 2026, the European AI Act — Regulation (EU) 2024/1689 — becomes generally applicable across the European Union. It is the world's first comprehensive AI law, and it directly affects any business with a chatbot talking to customers.

If you run a small business with an AI assistant on your website or WhatsApp, here is what changes, what you need to do — and what, if you use ChatME, is already done. (Selling to EU customers from outside Europe? The law follows the *user*, so it likely applies to you too.)

The AI Act in 60 seconds

The European law classifies AI systems into four risk levels:

• Unacceptable risk — banned since February 2025: harmful manipulation, social scoring, emotion recognition at work, and similar practices.
• High risk — allowed under strict requirements: systems deciding on credit, hiring, education admissions, health…
• Limited risk — allowed with a transparency obligation. This is where chatbots live.
• Minimal risk — most everyday AI, with no specific requirements.

Key dates: the law entered into force on August 1, 2024; the prohibitions apply since February 2, 2025; general-purpose model rules since August 2025; and general application — including chatbot transparency — arrives on August 2, 2026.

The obligation that touches your chatbot: Article 50

For chatbots, the heart of the Regulation is simple and reasonable:

When a person interacts with an AI system, they must know they are talking to a machine.

In practice, your assistant must identify itself as AI to your customers: on your website, on your chat link, and also on channels like WhatsApp or Instagram, where no interface makes it obvious.

Non-compliance isn't cheap: fines reach up to €35 million or 7% of global turnover for prohibited practices and up to €15 million or 3% for other infringements, with proportionality for SMEs. The sensible way to never think about those numbers: be transparent from day one.

What your chatbot must NOT do (and never should have)

The Regulation also forbids using AI systems like a chatbot to decide in high-risk areas: granting or denying credit, screening job candidates, deciding education admissions, or issuing medical diagnoses. Your assistant can inform about prices, treatments, requirements or timelines — the decision always belongs to a professional. If your chatbot today answers, informs, captures leads and books appointments, you are exactly on the right side of the line.

With ChatME, transparency is built in

We built ChatME as a European company, and the Regulation's key obligation is baked into the product — with zero configuration:

• Always-on AI disclosure: every widget shows the "AI" indicator in the header and the "You're chatting with an AI assistant" notice at the start of the conversation. It cannot be switched off.
• WhatsApp and Instagram covered: the first reply of every messaging-channel conversation automatically identifies itself as an AI assistant.
• Clear roles: under the Regulation, ChatME is the provider of the system and you are the deployer — with high-risk uses expressly excluded in our terms.
• EU data, EU provider: European infrastructure and GDPR compliance from day one.

Full detail on our security & compliance page and in the practical help-center guide.

Checklist for August 2

1. ✅ Does your chatbot identify itself as AI on your website? *(With ChatME: yes, by default.)*
2. ✅ And on WhatsApp/Instagram? *(With ChatME: yes, automatic.)*
3. 🤝 Do you use the assistant only to inform, capture and book — never to decide credit, hiring or diagnoses?
4. 🤝 Is the content you trained your assistant on up to date?
5. 🤝 Does your team know what the assistant does (and doesn't do)? — the "AI literacy" the law expects since 2025.

Five yeses, and August 2 is a day like any other.

A law that rewards doing it right

Transparency isn't just an obligation — it's trust. A customer who knows they're talking to AI and still gets their answer in seconds is a happy customer, and the businesses that comply from day one turn regulation into an advantage over tools that lag behind.

No assistant yet, or yours doesn't comply? Try ChatME free — in 10 minutes you have an assistant trained on your business and AI-Act-compliant from the first message.