Does your AI have to say it is an AI?

1. What three state laws say

Maine is the strictest of the three. Its statute (10 M.R.S. § 1500-DD) says a person may not use an AI chatbot "or any other computer technology" to do business with a consumer "in a manner that may mislead or deceive a reasonable consumer into believing that the consumer is engaging with a human being", unless the consumer is told "in a clear and conspicuous manner". The definition of a chatbot covers "textual or aural communications", so by its words it reaches a voice on the phone as well as a chat window. The law firm Cooley wrote on October 21, 2025 that the law took effect on September 24, 2025.

Utah asks for less. Under Utah Code 13-77-103 a business that uses generative AI with a consumer must say it is AI and not a human "if the individual asks", and the question has to be "a clear and unambiguous request". The definition covers text and audio. Cooley notes that before a 2025 amendment Utah required the notice at the start of every conversation. A stricter rule remains for occupations licensed through the Utah Department of Commerce: they must "prominently disclose" the AI at the start, but only in a "high-risk" interaction, which the statute describes as collecting health, financial or biometric data or giving financial, legal or medical advice. Whether an assistant that talks about financing or takes payment details falls there is a question for a Utah attorney.

California's law is narrower. Business and Professions Code § 17941 makes it unlawful to use a bot to communicate with a person in California online "with the intent to mislead the other person about its artificial identity" in order to push a purchase or sale. A person is not liable "if the person discloses that it is a bot". The statute defines "online" as a public-facing website or app; it does not name phone calls.

These are three laws, not a survey of fifty states. Other states have their own rules and we did not read them for this note.

2. What it costs

Saying it costs nothing: one sentence in the greeting.

Not saying it has a listed price. In Utah the consumer protection division may fine up to $2,500 for each violation, a court may impose the same, and breaking an order issued over a violation costs up to $5,000 each time (Utah Code 13-77-105). In Maine a violation is a violation of the Unfair Trade Practices Act; under 5 M.R.S. § 209 the Attorney General may seek a civil penalty of not more than $10,000 for each intentional violation. Cooley adds that the Maine act has its own limited right for a consumer to sue.

Utah also gives a safe harbor (13-77-104): no enforcement action for a missed disclosure if the AI itself clearly says "at the outset" and "throughout the interaction" that it is generative AI, is not human, or is an AI assistant.

3. Verdict

Worth adopting now: have the assistant say in its first sentence that it is an automated assistant, and have it answer honestly when someone asks. It is free, and it is the wording all three laws we read point to. We found no independent study of whether callers hang up more often after hearing it, so we give no number for that. This is not legal advice: check with your attorney which rules apply where your customers live.

Where WorkTime CRM fits, and where it does not

WorkTime CRM has a phone assistant that runs on your own Vapi and Twilio accounts, which you pay separately. The starting greeting the CRM gives it is: "this is Katie, your scheduling assistant, on a recorded line". That covers the recording notice. It does not contain the words "AI" or "automated", and the starting prompt does not tell the assistant what to answer when a caller asks whether it is a real person.

Both are yours to change. Open Settings → Connections → Vapi, press "Open Katie in Vapi" and edit the greeting and the prompt there. Then make a test call and listen. Keep in mind that the "✨ Use the Work Time prompt" button replaces your own prompt edits. WorkTime CRM does not know your state's rules and does not check them for you.

This week

1. 01Call your own number after hours and write down the assistant's first sentence word for word.
2. 02Ask it: "Am I talking to a real person?" Note what it answers.
3. 03Add "automated assistant" to the greeting, and one line to the prompt: when asked, say you are an AI assistant.
4. 04Do the same for a website chat or any automatic first reply: the first message says who is writing.
5. 05Ask your attorney one question: which AI disclosure rules apply in the states where my customers live?