Do landlords have to tell tenants they use AI?
In most states there is no general rule that a landlord must say a message was drafted with AI, but some states are starting to require disclosure in certain cases. Utah, for example, requires a business to disclose that a person is dealing with generative AI when the person clearly asks. The safest habit is simple: never deny it when asked, and review every message you send.
The short answer
There is no single federal rule saying a landlord must label AI-drafted emails or texts. Laws about AI are arriving state by state and mostly cover chatbots that talk to the public, not a landlord who uses an assistant to draft a note and then sends it. Whether a rule applies to you depends on your state and on how the AI is used. Check your state's current law, and treat anything here as general information, not legal advice.
One example: Utah
Utah's Code 13-77-103 says a supplier that uses generative AI to interact with someone in a consumer transaction must disclose that the person is dealing with AI rather than a human, when the person asks or otherwise prompts about it. The prompt must be a clear request to find out whether the interaction is with a human or an AI. Under that subsection the supplier has no duty to volunteer the disclosure unprompted.
Whether a given landlord-tenant exchange counts as a consumer transaction under Utah law is a legal question for your lawyer. The useful lesson is the pattern: if someone asks sincerely, answer truthfully.
Two different ways landlords use AI
| Use | Who is talking | Risk |
|---|---|---|
| AI drafts, you review and send | You | Low. The message is yours. Your name is on it. |
| AI answers tenants on its own | An automated assistant | Higher. Disclosure rules and mistakes are more likely to matter. |
If you review and send every message, you are the author. You are responsible for what it says, which is why a landlord is responsible when AI makes a mistake. If an assistant talks to tenants without you, read the risks of letting AI answer tenant messages first.
Habits that keep you safe whatever the law says
- If a tenant asks whether a message was written by AI, say yes plainly.
- Read every message before it goes out and check names, amounts and dates.
- Do not let an assistant make legal threats, set deadlines or discuss anyone's protected traits.
- Keep a record of what was sent and when.
- Offer a way to reach you directly for anything urgent.
Example a made-up tenant replies to a maintenance text, "Is this a robot?" The owner answers, "I use an AI tool to draft my messages and I read each one before sending. You can call me at this number any time." That is honest, short and keeps the relationship.
Should you add a label anyway?
A line in your lease welcome note such as "I use an AI tool to help draft some messages; I review everything I send" costs nothing and heads off surprise. It also removes the temptation to hide it. It is your choice. If you use an assistant that replies by itself, a clear label at the start of the exchange is the cautious path.
How DoorKeeping handles it
With DoorKeeping, your assistant shows you a draft first and sends one message to one tenant only after you confirm, so the message is yours. Tenants who opted out of texts get nothing. Rent stays in your own bank account, and anything involving money or a legal step waits for your approval. DoorKeeping does not decide whether a disclosure law applies to you; check your state or ask a lawyer.
Sources
- Utah Code 13-77-103: Required disclosures (generative AI), read Oct 31, 2026
- DoorKeeping: Connect Claude or ChatGPT (steps and tools list), read Oct 31, 2026
Questions
- Is it illegal to use ChatGPT to write messages to tenants?
- No general rule bans it. You remain responsible for what you send, so review every message.
- Must I tell a tenant a chatbot is not a person?
- In some states, such as Utah, a business must disclose AI use when a person clearly asks. Rules vary and are changing, so check your state.
- What if a tenant asks whether it was AI?
- Say yes plainly. Never deny it.
- Does it matter if I review the message first?
- Yes. If you review and send it, you are the author and responsible for it, and the disclosure rules aimed at automated chat usually matter less.