Your chatbot isn't the only thing that owes a disclosure under Article 50.
Almost every conversation I have about Article 50's disclosure duty starts from the same assumption: it's a chatbot rule. Someone asks whether their support widget needs a disclosure notice, we talk it through, and that's the whole scope of the conversation. It's a reasonable assumption, since chatbots are the loudest, most visible case. It's also not the full rule.
The disclosure obligation covers three categories of systems, and two of them have nothing to do with conversation at all: emotion recognition systems, and biometric categorization systems. If your product infers how someone is feeling from their voice, expression, or behavior, or sorts people into categories based on biometric data, you're carrying the same disclosure duty as a chatbot team, even though nothing about what you built looks like a chatbot.
It's worth being precise about what each one actually covers. An emotion recognition system is anything that infers or classifies a person's emotional state from their behavior, expressions, or voice — a call-center tool scoring customer frustration, a video interview platform flagging candidate confidence, a camera system reading shopper mood. A biometric categorization system is anything that sorts people into categories based on biometric data, separate from standard biometric identification or verification (unlocking a phone with your face isn't this; inferring someone's age bracket or demographic group from a camera feed is). Neither of these has to talk to anyone to trigger the duty.
That's exactly why they get missed. A chatbot team knows it's building something that interacts with people, so the “does this need a disclosure” question at least gets asked. A sentiment-scoring feature bolted onto a call-center analytics dashboard, or a demographic-inference module inside a retail camera product, usually isn't framed as an “AI system that talks to people” internally at all — it's framed as an analytics feature. The disclosure duty doesn't care how the feature is framed internally. It cares whether a person is being analyzed by it.
There's a subtlety worth flagging here too: the disclosure isn't owed to whoever is looking at the dashboard. It's owed to the people being analyzed. For a call-center tool, that means the caller whose tone is being scored, not the agent or the manager reviewing the report. For a retail camera system, that means the shopper being categorized, not the store operator who bought the product. Teams that do think about disclosure for these systems often think about it from the buyer's side — informing the business customer what the product does — which is a good practice, but it isn't the disclosure Article 50 is asking for.
Everything else about the duty carries over unchanged from the chatbot case. It's a disclosure requirement, not a consent requirement — the person needs to be informed, not asked to opt in. The “obvious from context” exception applies the same way: obvious to the person being analyzed, not to the team that built the system. And it isn't the only obligation in play, either — being able to show that the disclosure actually happened is a separate audit-trail duty, exactly as it is for a chatbot interaction.
That last part is why Disclarion's SDK logs a disclosure event with the jurisdiction and interaction type attached, rather than assuming every interaction is a chat message. An emotion-recognition scoring call and a chatbot reply aren't the same kind of interaction, but they owe the same two things underneath: a person informed, and a record that they were. Coverage can be checked against whichever obligation actually applies to a given system, instead of everything getting funneled through chatbot-shaped assumptions.
If your product touches emotion inference or biometric categorization anywhere, even as a secondary feature inside a larger tool, it's worth checking against the full scope rather than the chatbot-only version of this rule most teams start from. For the complete breakdown of which system types trigger disclosure and what specifically has to be shown, see our disclosure requirements guide.
More from the blog
Your internal AI tools owe the same disclosure duty the customer-facing one does
An HR bot or IT helpdesk assistant used only by your own staff still interacts with natural persons. Article 50 doesn't carve out an exception for "internal-only."
Does the EU AI Act apply if you're not based in the EU?
"We're a US company, we don't have an EU entity" isn't the exemption it sounds like. The actual test is about who's on the other end of the conversation, not your mailing address.
Why I built Disclarion as an SDK, not a dashboard
The easier product to build is a dashboard you configure after the fact. I built Disclarion to sit inside the AI call itself instead, and it's worth explaining why.
Deepfakes get a stricter rule than an AI-written blog post
Article 50(4)'s deepfake duty isn't the same obligation as ordinary AI content labeling — it applies to image, audio, and video specifically, and the built-in exemptions work differently too.
You named it "AI Assistant." That's not the same thing as disclosing it.
A bot's name feels like disclosure. Article 50 cares whether a reasonable user actually understood they were talking to an AI — not whether a label existed somewhere on the page.
Your editor touched it. Article 50's label exemption still might not apply.
Article 50(4) waives the visible AI-content label when a human takes substantive editorial responsibility for a piece — but a quick skim doesn't qualify, and the exemption doesn't touch the audit-trail duty underneath it.
You labeled the article. That's not the same thing as marking it.
Article 50(2)'s machine-readable marking and Article 50(4)'s visible label are two different duties. Images already meet the first one by accident. Article text usually doesn't meet it at all.
Your AI answers the phone now. There's no modal for that.
Article 50's disclosure duty doesn't go away because there's no screen. What actually discloses an AI voice agent on a phone call, when a modal isn't an option.
An AI audit trail doesn't need to know what your users actually said.
Teams hear "audit trail" and picture a database of every conversation. Article 50's version is narrower: it has to prove disclosure and labeling happened, not repeat what was said.
Article 50 skips disclosure when it's "obvious from context." That word is doing more work than it looks like.
Disclosure isn't required when AI use is obvious from context, but Article 50 never defines obvious, and most of the cases teams assume qualify, don't.
Your chatbot needs to disclose it's an AI. Once, not every message.
Disclosing a chatbot is AI sounds simple until you build it: repeat it every message and people tune it out, show it once and later visitors never see it.
You have logs. That's not the same thing as an audit trail.
Under Article 50, an audit trail has to be insert-only, complete, and producible on demand, not a data-engineering project when a regulator actually asks.
Your AI-generated content needs a label. Here's what that actually means.
A disclaimer someone remembers to add isn't a labeling system. Under Article 50, AI-generated content needs to be identifiable as such, reliably, not as a manual step someone can forget.
EU AI Act Article 50, explained: who it affects and what it requires
Article 50 of the EU AI Act has been enforceable since August 2, 2026: who it applies to, its three obligations, and the penalties for getting it wrong.