EU AI Act Article 50: when you must disclose that someone is talking to AI

A common assumption about AI tools for social media and inboxes: "It's just an auto-reply — the recipient doesn't need to know it's AI." Under the EU AI Act, that's not how it works.

What Article 50(1) says

Any AI system intended to interact directly with a person must be designed so the person is informed they're interacting with AI — unless that's obvious from the context.

The distinction is practical. An AI auto-reply on LinkedIn that mimics your personal voice? Not obvious — its whole purpose is to sound like you. Disclosure applies. A classic out-of-office "I'm away until Monday"? Obvious — the format itself signals automation. No disclosure required.

Why it matters

The fix is tiny: one line, at the start or end — "This reply was sent with the help of AI." But its absence is no longer a stylistic choice; it's a regulatory gap. Most popular automation tools don't do this by default. If you run one, it's worth checking before someone else does.

Article 50's transparency connects to the rest of the regulation — record-keeping and the core obligations for small businesses — and to reliability: a bot that invents answers without even disclosing it's AI exposes you twice.

What to do

• Map every point where AI replies to humans (chatbot, auto-reply, voice agent).
• Add a clear, unobtrusive AI disclosure wherever it isn't obvious.
• Document the decision — where you added disclosure and why.