Do You Have to Tell Callers Your AI Receptionist Isn’t Human? What the New State Laws Say

Customer service representative wearing a headset while answering a call at a call center desk

Short answer: in a growing number of states, yes, at least in some situations. Maine already bans using a chatbot in a way that could make a reasonable consumer think they are dealing with a human, unless you clearly tell them it is not. Utah requires you to disclose generative AI when a customer asks. California signed a “Right to Human Customer Service” law in late September 2026 that targets very large companies. For a small HVAC, plumbing or property management business, the simplest rule is to have your AI say it is an AI at the start of every call and chat, and to give callers an easy way to reach a person.

Key Takeaways

  • Maine law (10 M.R.S. §1500-DD) covers chatbots that talk by text or voice. If a reasonable consumer could think they are talking to a human, you need a clear and conspicuous notice that they are not.
  • Utah requires disclosure when a consumer clearly asks whether they are talking to AI. Businesses that disclose at the outset and throughout the conversation get a safe harbor from enforcement.
  • California’s AB 1609 applies only to businesses with more than $500 million in national annual revenue. It still shows where expectations are heading: disclose the AI and offer a fast path to a human.
  • Colorado’s HB26-1263 adds disclosure duties for “conversational artificial intelligence services” starting January 1, 2027. Ask a lawyer whether it reaches a business phone assistant.
  • A short disclosure line plus a reliable transfer to a person costs you almost nothing. It protects trust and covers the strictest rules you are likely to face.

Picture a tenant calling at 10 p.m. about a leaking water heater. A friendly voice answers, takes the unit number and books a plumber for the morning. A week later the tenant mentions the “nice woman on the phone,” and your office manager has to explain there was no woman. Most customers shrug that off. Some feel misled. In several states, regulators now care about that moment too.

We have covered how the TCPA treats AI voice on outbound calls. This post covers the other side: inbound calls and chats your AI answers, and what new state laws say about telling people it is AI.

What Do the New AI Disclosure Laws Actually Say?

No single national rule covers this. The laws we reviewed are state laws, and they differ in who they cover and when disclosure is required. Here is what each primary source or law-firm summary says.

Maine: don’t let a chatbot pass as a human

Maine’s statute defines an “artificial intelligence chatbot” as a software application, web interface or computer program that simulates human conversation through text or voice. That covers an AI phone receptionist as well as a website chat widget.

The rule says you may not use a chatbot in trade or commerce in a way that may mislead or deceive a reasonable consumer into believing they are engaging with a human, unless the consumer gets a clear and conspicuous notice that they are not. A violation is treated as a violation of the Maine Unfair Trade Practices Act.

Note what triggers it. The customer does not have to ask. If your AI sounds human and nothing tells the caller otherwise, you are exposed.

Utah: disclose when asked, and get a safe harbor for disclosing up front

Utah amended its AI law in 2025 with SB 226, effective May 7, 2025. According to a Davis Polk client update:

  • A business using generative AI to interact with consumers must disclose it when the consumer makes a “clear and unambiguous request” to know whether they are talking to a human or AI.
  • Prominent disclosure at the outset is required for “high-risk” interactions. These involve collecting sensitive information, such as financial, health or biometric data, and giving personalized advice people may rely on for major decisions.
  • If you clearly and conspicuously disclose that the consumer is interacting with generative AI, both at the outset and throughout the interaction, you get a safe harbor from enforcement.
  • Routine AI-generated appointment confirmations and reminders sent by text are excluded from the AI system definition.
  • The Division of Consumer Protection can impose administrative fines of $2,500 per violation. There is no private right of action. The law’s repeal date was extended to July 2027.

California: the Right to Human Customer Service Act (AB 1609)

Governor Newsom signed AB 1609 in late September 2026, along with more than 20 other AI-related bills, according to Covington’s Inside Privacy. Quarles notes that the standard effective date for such a law would be January 1, 2027.

According to a Fox Rothschild summary dated September 30, 2026:

  • It covers “large private businesses” with more than $500 million in gross annual revenue nationally.
  • Covered businesses must clearly and conspicuously disclose that a customer service chatbot is not human when a reasonable person could be misled. The disclosure must be in the same medium as the conversation.
  • During regular business hours, customers need a simple way to ask for a human. The business must make a good-faith effort to connect them within 15 minutes or schedule a specific appointment within one business day.
  • Civil penalties run up to $5,000 for a first violation and up to $10,000 for each later one, enforced by a public prosecutor.

Almost no local contractor or property manager is anywhere near $500 million in revenue. Still, this law signals where customer expectations are heading.

Colorado: conversational AI operators, starting 2027

Colorado’s HB26-1263 was signed May 29, 2026, and its operator requirements take effect January 1, 2027. Operators of conversational AI services must tell users the service is artificial intelligence, and the law adds extra protections for minors. The bill summary describes covered services as AI systems available to the general public that mainly simulate human conversation. It does not clearly say whether a business’s own receptionist bot counts. Ask your attorney if you operate in Colorado.

Why Does This Matter for a Small Home Service or Property Management Business?

These laws target the moment when a customer is misled about who they are talking to. AI receptionists are built to sound natural, which is why they work, and also why the question comes up.

Three practical reasons to care:

  1. Your customers may live in more than one state. Property managers with portfolios across state lines and contractors near a border can easily end up serving a Maine or Utah resident.
  2. Trust is the product. A homeowner letting a technician into their house, or a tenant reporting a maintenance problem, wants to feel dealt with honestly. Finding out later that the “person” was software can sour the relationship more than an upfront disclosure ever would.
  3. The rules are spreading. Several states acted in just two years. Building disclosure and human handoff into your setup now is easier than re-recording greetings and retraining staff later.

What Does Good AI Disclosure Look Like in Practice?

You do not need a legal paragraph. You need one honest sentence early, and a real way out.

For phone calls

Open with something like: “Thanks for calling Smith Heating & Air. I’m the company’s AI assistant. I can book a visit, take a message or connect you with our team.” That one line tells the caller it is AI, sets expectations and offers a human.

For texts and website chat

Put the disclosure in the same channel. California’s law, for example, requires it in the same medium as the conversation. A first message like “Hi, this is the AI assistant for Oak Street Property Management. A team member can step in anytime” works for both a website chat and a missed-call text.

Industry examples

  • HVAC, no-heat call at night: The AI says it is an assistant, collects the address and symptoms, and routes the call to the on-call tech. It does not troubleshoot the furnace.
  • Plumbing, burst pipe: The AI tells the caller how to reach the on-call plumber and directs anyone describing a gas smell or safety hazard to emergency services. Your AI should collect details and route the call, not give repair or safety advice.
  • Property management, rental lead: The AI chat agent answers questions about listed units from your approved information, books a showing and hands off to a leasing agent for anything about application decisions.
  • Real estate, buyer inquiry: The AI qualifies the lead with your chosen questions and books a call with an agent. It does not give pricing or financial advice.

Comparing the Rules at a Glance

State / law Who it covers When disclosure is required Timing
Maine, 10 M.R.S. §1500-DD Anyone using a chatbot (text or voice) in trade or commerce with consumers Whenever a reasonable consumer could be misled into thinking it is a human In effect
Utah, SB 226 amendments Businesses using generative AI with consumers When the consumer clearly asks; at the outset for high-risk interactions Effective May 7, 2025; repeal date extended to July 2027
California, AB 1609 Businesses with more than $500M national revenue When a reasonable person could be misled; plus human access within set time frames Signed late September 2026; expected January 1, 2027
Colorado, HB26-1263 Operators of conversational AI services Must tell users the service is AI; extra rules for minors Operator requirements start January 1, 2027

This table summarizes the sources linked in this post. It is not legal advice. Confirm how these laws apply to your business with a qualified attorney.

Benefits and Challenges of Disclosing Up Front

Benefits

  • One standard for every state. Disclosing on every call meets Maine’s rule and Utah’s safe harbor approach at the same time.
  • Fewer awkward moments. Callers who know it is AI speak more clearly and don’t ask it things only a person can answer.
  • Cleaner handoffs. Offering a human at the start means frustrated callers ask for one instead of hanging up.

Challenges

  • Some callers will hang up. A few people refuse to talk to software. That is why a working transfer and a fast callback matter.
  • Your human backup has to be real. Promising “our team can help” and then sending everyone to a full voicemail box defeats the purpose.
  • Rules keep changing. Utah’s law has a repeal date, Colorado’s rules start in 2027 and other states are active. Review your setup at least once a year.

How Should You Set Up AI Disclosure? Six Steps

  1. List where your AI talks to customers. Phone, missed-call texts, website chat, Facebook, Instagram, WhatsApp. Each channel needs its own disclosure.
  2. Write one disclosure line per channel. Keep it short: business name, “AI assistant,” and what it can do.
  3. Put it first. Disclosure at the start is what earns Utah’s safe harbor and keeps you clear of Maine’s “misled” standard. Don’t bury it after the booking questions.
  4. Answer the question honestly every time. If someone asks “Am I talking to a real person?” the AI should say no. Test this yourself.
  5. Set up the human path. Choose who gets transferred calls, during which hours, and what happens after hours. Make test calls to confirm it works.
  6. Keep sensitive topics with people. Rent disputes, billing complaints, fair housing questions and anything involving safety should go to a staff member.

Common mistakes to avoid

  • Giving the AI a human name and backstory (“This is Sarah, I’ve been with the company ten years”).
  • Disclosing on the phone but not in the follow-up text or chat.
  • Transfers that ring a phone nobody answers.
  • Letting the AI give repair, legal or safety advice instead of collecting details and routing.
  • Assuming that because California’s law targets big companies, no rule applies to you.

EngagePro’s Perspective

We build EngagePro for owners who need calls and messages answered while the team is on a job. Disclosure and human handoff are part of setup, not an afterthought. Here is how we’d set it up in EngagePro.

Start with the transfer, not the script. The EngagePro AI Receptionist can answer during business hours, after hours, weekends and holidays based on your settings, and can transfer the caller to a person or phone number you choose. Decide who that person is for each time window. Then place a live test call and confirm the transfer goes through before you turn the receptionist on for real customers.

Tell callers it’s AI, in your words. Whatever tool you use, check that the first thing callers hear identifies the assistant as AI. Then call your own line and ask, “Am I talking to a real person?” If the answer isn’t a clear no, fix it before going live.

Use the same standard for text and chat. The AI Text & Chat Agent responds on SMS, Facebook, Instagram, WhatsApp and website chat. You can set topics it should avoid and the conditions for handing off to a person, and your team can step in from the unified inbox whenever a conversation needs it. Put disclosure in the first message, and add rent disputes, complaints and safety issues to the hand-off list.

Keep the record. Caller details are saved to the contact record in your CRM, and the receptionist can send a text or start a workflow based on what happened on the call. If a customer later says they didn’t know they were talking to AI, you can look at the conversation history and fix the gap.

One limit to be clear about: EngagePro is a tool. It does not make your business compliant with any law. You decide what your AI says and where it routes people, and you should confirm the rules for your states with a qualified attorney.

FAQ

Is there a federal law that requires my AI receptionist to say it’s AI? The disclosure laws covered in this post are state laws: Maine, Utah, California and Colorado. Federal rules on AI voice in outbound calls are a separate topic. Ask your attorney whether any federal rule applies to your inbound setup.

My business is in Texas. Do I need to worry about Maine’s law? Possibly, if you serve Maine consumers, for example through out-of-state rental properties or customers. Ask your attorney how these laws apply across state lines. Disclosing on every call avoids the guesswork.

Does California’s AB 1609 apply to my plumbing company? According to the law-firm summaries we reviewed, it covers businesses with more than $500 million in gross annual revenue nationally. Most local service businesses are far below that. The practices it requires, disclosure and access to a human, are still good habits.

Do appointment reminder texts need an AI disclosure? Under Utah’s amended law, routine AI-generated appointment confirmations and reminders sent by text are excluded from the AI system definition, per Davis Polk. Other states may differ. Separate texting rules on consent and opt-outs still apply.

Can I give my AI receptionist a name? A name is fine if the caller is also told it’s an AI assistant. The risk comes from presenting it as a human. Under Maine’s law, the test is whether a reasonable consumer could be misled.

Will disclosing that it’s AI hurt my booking rate? Some callers will ask for a person. Plan for that with a working transfer during business hours and a fast callback after hours, so those callers become bookings instead of hang-ups.

What should my AI never do, even with disclosure? Give repair, safety, legal or financial advice. It should collect details, book or route the call, and send true emergencies to emergency services.

How often should I review my AI setup? At least once a year, and whenever you expand into a new state. Colorado’s requirements start January 1, 2027, and Utah’s law has a July 2027 repeal date, so things will keep moving.

Conclusion

An AI receptionist is supposed to make sure no call goes unanswered while your crew is on a job. It shouldn’t leave customers wondering who they talked to. Maine, Utah, California and Colorado have each drawn a line around AI pretending to be human. The fix fits in one sentence: say it’s an AI assistant, and give people a real way to reach your team.

If you want help setting up AI call answering and texting with clear disclosure and reliable human handoff, start a free trial or see how EngagePro works for small businesses. Then confirm your state’s rules with your attorney before you go live.

Further Reading