Chatbot transparency: when the label matters
Users must be told they are interacting with AI unless it is obvious from context; the exam trap is assuming 'obvious' always applies to simple bots.
Слушай тази страница (бета)
The rule is straightforward: if a user interacts with an AI chatbot, they must be informed that it is AI, not a human. The only exception is when it is 'obvious from the context'—for example, a clearly labelled automated customer-service portal. The key contrast is between a chatbot that mimics human conversation (where transparency is mandatory) and a basic form-filler that no one would mistake for a person (where it may be waived).
A common exam trick is to present a simple, repetitive chatbot and claim the context makes AI obvious—but if the bot uses natural language or a human-like name, the exception likely does not apply. Another hint: the obligation is about the interaction itself, not the task. Even a bot that only answers FAQs must disclose its AI nature if it could be mistaken for a human. To eliminate wrong options, check whether the user could reasonably believe they are chatting with a person; if yes, disclosure is needed.
To test yourself, imagine a chatbot named 'Sarah' that answers support queries. Ask: would a user assume Sarah is human? If yes, the transparency rule applies regardless of how simple the bot is. This mental check cuts through the 'obvious from context' exception and keeps you focused on user perception.
What must users generally know when interacting with an AI chatbot?
That they are interacting with AI, unless this is obvious from the context.