Chatbot disclosure

What California law says

California law bars using a bot to deceive a person there about whether they are dealing with a machine, when the aim is to encourage a sale. Disclosing the bot takes it outside that rule, and the disclosure has to be clear and conspicuous (California Business and Professions Code).

The rule turns on misleading people, so the remedy is to be plain about what the bot is. This page is general information, not legal advice.

What Utah law says

Utah's current disclosure rules for generative AI in consumer transactions have been in effect since May 7, 2025. A business that uses generative AI to talk with a person in a consumer transaction has to tell them they are dealing with AI and not a human if they clearly ask (Utah Code, Title 13, Chapter 77).

People in licensed occupations carry a stricter duty. When a high-risk interaction, such as collecting health data or giving legal or medical advice, runs through generative AI, they have to disclose it prominently at the start. Utah also treats disclosure made clearly and conspicuously at the outset, and kept up throughout the interaction, as a safe harbor.

Saying it plainly at the start of every chat

So in California and Utah, in the situations above, the law asks for it. Anywhere else, check your own state's rules, and disclose on every chat regardless.

Put it in the first message, in words, in the same size of text as the rest of the chat. Something like: "Hi, I am an automated assistant, not a person. I can answer questions about our services and hours, or pass you to the team."

Then keep it true for the whole conversation. If a visitor asks whether they are talking to a human, the bot answers no and offers a person. Do not give it a staff member's name or photo, and do not script lines such as "let me check with my manager" that suggest someone is typing.

How our chatbot handles it

The AI chatbot we offer answers, qualifies and hands off, and always says it is a bot. It says so when each conversation opens. Asked whether a person is on the other end, it answers that it is a bot and offers to pass the visitor to your team, by the phone number, booking link or email you picked.

The free scan cannot check a chat window for you. It reads your pages as your server sends them and does not run their scripts, so a chat widget that loads by script is outside what it sees (what the scan does not measure).

Questions

Is a small AI label in the chat window enough?

We would not rely on one. California asks for a disclosure that is clear and conspicuous, and Utah's safe harbor asks for one at the outset and throughout the conversation. A small badge in a corner is easy to miss on a phone. Say it in words in the first message, and keep the label as well.

Do I need to disclose a bot that only books appointments?

Disclose it anyway. California's rule is written around sales, and Utah's around consumer transactions and licensed occupations. Whether a booking assistant falls inside either is a question for a lawyer who knows your situation. Saying it is a bot costs one sentence in the first message.

Where this fits in our work

We handle this as part of AI chatbot: A chatbot trained on your own site that answers, qualifies and hands off to your team, and always says it is a bot.

Related reading

Related terms

• AI hallucination: When an AI tool states something false with confidence, why it happens, and what it means for a chatbot that speaks for your business.
• Business associate agreement (BAA): The HIPAA contract a covered health practice signs with a vendor that handles its patient information, such as the company behind a patient chatbot.

Sources

• California Legislative Information: Business and Professions Code, sections 17940 to 17943. Checked September 24, 2026.
• Utah Legislature: Utah Code Title 13, Chapter 77: Generative Artificial Intelligence, Consumer Disclosures and Enforcement. Checked September 26, 2026.