---
title: "Airbnb AI disclosure rules for hosts: what Airbnb, Booking.com, the EU AI Act and US state laws require"
description: "Airbnb has no explicit rule making hosts tell guests a reply came from AI, and neither does Booking.com. The EU AI Act requires it from 2 August 2026 when AI sends on its own. California and Utah add rules. Wording to paste included."
url: https://supalobby.com/en/blog/airbnb-ai-disclosure-rules-for-hosts
language: en
published: 2026-10-20
updated: 2026-10-20
author: Bartosz Wilczyński
translation: https://supalobby.com/blog/czy-musisz-mowic-gosciom-ze-odpisuje-ai
---

# Airbnb AI disclosure rules for hosts: what Airbnb, Booking.com, the EU AI Act and US state laws require

Airbnb has no explicit rule making hosts tell guests a reply came from AI, and neither does Booking.com. The EU AI Act requires it from 2 August 2026 when AI sends on its own. California and Utah add rules. Wording to paste included.

Airbnb has **no explicit rule that makes hosts tell guests a message was written by AI**, and neither does Booking.com (checked 26 September 2026). The EU AI Act requires disclosure from 2 August 2026 when the AI sends replies on its own rather than drafting them for you. California and Utah add their own rules.

Below: what the platforms' pages say, what Article 50 of the AI Act requires and from whom, six ways hosts use AI and whether each needs a notice, two US state laws, and wording to paste.

## What Airbnb's own rules say, and do not say

Airbnb has no explicit host-facing rule on disclosing AI-written messages. We read the Content Policy, the Terms of Service, the page on AI-powered features and the Resource Center on 26 September 2026.

The Content Policy defines content as „any written, photographic, audio, video, or other content, including any content generated or edited using AI”, and lists messages to guests among the things it covers. Its bans cover illegal content, harassment, impersonation, advertising a separate business and similar. There is no labelling requirement.

The Terms of Service (effective 5 February 2026 for new members, 20 April 2026 for existing ones) say in section 11.1: „Do not use bots, crawlers, scrapers, or other automated means to access or collect data or other content from or otherwise interact with the Airbnb Platform.” The practical reading: do not use browser plug-ins that click around your Airbnb inbox on their own, and the safest set-up is that you press send on Airbnb yourself. Section 9 lets Airbnb offer its own tools to generate or translate content. The 2026 update added that evidence in Host Damage Protection claims must not include AI-generated content.

Airbnb labels its own AI. The Resource Center piece of 20 May 2026 describes auto-replies in the Messages tab: they answer common questions with details from your listing, each one identifies Airbnb as the sender, both you and the guest can see it, and you can delete it and write your own. AI-suggested quick replies work the other way: only you see them, and you review and edit before sending.

Secondary sources agree. 404 Media reported on 14 April 2026 that Airbnb permits certain hosts to use tools that reply on their behalf outside their usual hours; it quotes no disclosure rule. Hostrexa wrote on 20 March 2026 that Airbnb's terms do not require hosts to disclose AI-assisted replies, without linking an Airbnb page. HostReply, a vendor of AI reply tools, is sometimes cited for „Airbnb AI-disclosure rules”; its blog had no such piece when we checked on 26 September 2026.

## What Booking.com's rules say

Booking.com has no explicit AI disclosure rule for partners either; we found none in the Partner Hub on 26 September 2026. It has its own AI tools and one clear instruction: communicate with guests only through Booking.com's platforms, the Extranet and the Pulse app.

„Help me reply” is a button at the bottom of the chat window in the Extranet, also on the reservation details page. It suggests a reply, and the instruction reads: review the suggested message carefully, you can edit it and, if you agree, send it. That is a drafting tool, indirect interaction in the Commission's terms. You switch it on or off with the three dots in the top right corner of the chat window.

Automatic replies work differently: you set them once and they go out on their own. The guest receives them by e-mail and sees them in the chat. Booking.com's messaging tools page says over 30 percent of all responses from partners are automatic replies. On 9 October 2025 Booking announced Auto-Reply, available globally, and Smart Messenger, which gathers partner, property and reservation information to „autonomously take action”, knowing when and what to suggest. Smart Messenger is live for partners based in the US, UK, New Zealand and Australia and those who chose English. None of these pages says whether the guest sees a label on automatic replies, and none requires partners to add one.

For an AI agent in the Booking.com thread, the disclosure line has to fit in that thread, and the rule is the same as elsewhere: a suggestion you send yourself needs nothing; a reply that leaves on its own needs one sentence about AI.

## What the EU AI Act requires from 2 August 2026

Article 50(1) obliges the provider of an AI system intended to interact directly with people to design it so that those people are informed they are interacting with an AI system, unless that is „obvious from the point of view of a natural person who is reasonably well-informed, observant and circumspect”. Article 50(5) adds that the information must be given „in a clear and distinguishable manner at the latest at the time of the first interaction or exposure”.

Who is who:

- The **provider** develops an AI system and places it on the market under its own name or trademark (Article 3(3)). That is the maker of the tool you use.
- The **deployer** is anyone „using an AI system under its authority except where the AI system is used in the course of a personal non-professional activity” (Article 3(4)). A host who rents for money and switches an AI tool into guest conversations is hard to call a private user.

The letter of the law: the Article 50(1) duty sits with the provider, not the host. The Commission's guidelines of 20 July 2026 say the provider must make sure every person interacting with the system is informed, irrespective of who deploys it (paragraph 11). In practice, though, the notice appears in your thread with the guest, and whether a reply leaves without a human is your setting. The guidelines also close the loophole: the mere possibility for humans to intervene or review outputs should not be used to circumvent the disclosure requirement (paragraph 30).

Dates and penalties. The regulation applies from 2 August 2026 (Article 113), and the Commission's Q&A confirms Article 50 applies from that date. Non-compliance with Article 50 carries a fine of up to 15 million euro or 3 percent of worldwide annual turnover, whichever is higher (Article 99(4)); for SMEs the lower of the two applies (Article 99(6)). Each member state names the authority that enforces it.

## When a disclosure is on you as a host

In practice, when AI sends the guest a reply without you. When AI writes a draft that you read and send, no notice is needed. That is where the Commission's guidelines draw the line.

| How you use AI | Is the guest interacting with AI directly | What to do |
|---|---|---|
| AI suggests a reply, you read and send (Copilot mode, Booking's „Help me reply”, Airbnb's AI-suggested quick replies) | no; guidelines, paragraph 30: customer service staff using AI assistance tools is mediated interaction | nothing |
| Templates and scheduled messages with no AI, out-of-office replies | no; paragraph 30 excludes simple non-AI automated response mechanisms | nothing |
| AI answers on its own on your website chat, WhatsApp, Instagram or in the Booking.com thread (Autopilot mode) | yes | one sentence about AI in the first message |
| AI answers the property phone | yes | a spoken notice at the start of the call |
| Some replies by AI alone, some by you | yes for the AI replies, unless a human properly reviewed and sent them as the main interlocutor (paragraph 30) | a notice at the start of the thread |
| An AI agent that books, changes dates or answers inquiries on your behalf | yes; paragraph 31: agents must disclose their artificial nature and the person on whose behalf they act | put the property's name in the notice |

The „obvious from context” exemption is not for you. The guidelines read it narrowly (paragraph 45): general awareness that bots exist is not enough, and the worked examples where it applies concern professionals, such as a code assistant for developers or an internal assistant for trained staff. A guest is a consumer, sometimes tired after a journey, sometimes older and less used to AI. Assume you must tell them.

The form is simple. The guidelines give as examples a chatbot that starts the conversation by mentioning it is based on AI, an e-mail from an AI agent with an AI label at the top, and a voice assistant that says it is powered by AI at the beginning of a session (paragraph 36). A single prominent notice before the first interaction is likely to suffice in most cases (paragraph 40). What does not suffice on its own: a line only in terms and conditions or documentation, an ambiguous signal such as a generic „assistant”, a human-like persona that may mislead, or a blanket „services on this website use AI” (paragraph 38).

## What US state laws add

For US hosts, two state statutes bear on AI replying to guests.

California's Business and Professions Code, sections 17940 to 17943, operative since 1 July 2019, makes it unlawful to use a bot to communicate or interact with a person in California online „with the intent to mislead the other person about its artificial identity” in order to incentivise a purchase or sale of goods or services in a commercial transaction. A bot is „an automated online account where all or substantially all of the actions or posts of that account are not the result of a person”. Whether an AI replying from your host account meets that definition is open. The safe harbour: a person using a bot is not liable if they disclose that it is a bot, in a way that is clear, conspicuous and reasonably designed to inform. A one-line notice removes the question.

Utah's generative AI consumer protection law, as amended by SB 226 with effect from 7 May 2025, requires a business to disclose generative AI use when a consumer makes „a clear and unambiguous request” to know whether they are dealing with a human, according to a Davis Polk summary; we could not open the statute page itself. A safe harbour applies to businesses that disclose clearly and conspicuously at the outset and throughout the interaction. Prominent upfront disclosure is mandatory only for high-risk interactions that combine sensitive personal data with advice people rely on for significant decisions, which ordinary guest messaging is unlikely to be.

## Ready-to-paste disclosure wording

One sentence at the top of the first reply, with the word „AI” spelled out, the property's name and a way to reach a human. In the guest's language.

Website chat, WhatsApp, Instagram:

```
Hi, this is the AI assistant for Lakeside Cabin. I answer straight away about directions, check-in and what is in the cabin. If you would rather talk to the host, type "human" and I will hand the conversation to Sarah.
```

Booking.com thread, shorter:

```
This reply comes from the AI assistant for Lakeside Cabin. Ask for the host at any time and I will pass the conversation on.
```

Phone greeting:

```
Hello, you have reached Lakeside Cabin. You are speaking with an AI assistant. I can check available dates and answer questions about your stay. Say "host" at any time to speak to a person.
```

The same for Polish-speaking guests:

```
Dzień dobry, tu asystent AI obiektu Lakeside Cabin. Od razu odpowiadam na pytania o dojazd, zameldowanie i wyposażenie. Jeśli wolisz rozmawiać z gospodarzem, napisz „człowiek”, a przekażę wątek Sarze.
```

Put it in the first message of every new thread, as a persistent line in your website chat header, and at the top of an e-mail, not the footer. Do not bury it in the house rules, give the agent a human name with no „AI” next to it, or write something vague like „we use modern technology”; the guidelines (paragraph 38) treat similar approaches as insufficient.

If the agent only drafts and you send, no notice is required. You may add one as a courtesy, but you do not have to.

## Where Supalobby fits

Supalobby has two modes, and they decide whether the notice is needed. In Copilot the agent drafts every reply and you send it. On Airbnb it only works this way: it reads the thread and drafts, and you send with one click. In Autopilot the agent replies on its own, switched on per channel. You can keep e-mail in draft mode and run the website chat and WhatsApp in Autopilot with the AI notice in the first message. When the agent is unsure it asks you and hands the thread over, so the guest has a route to a person. The wording stays with you; Supalobby is not a law firm and does not review your privacy notice.

## Frequently asked questions

**Does Airbnb require hosts to disclose AI-written messages?**

We found no such rule on 26 September 2026. Airbnb's Content Policy covers content generated or edited using AI, including messages to guests, and applies the same bans as to anything typed by hand. The Terms of Service forbid bots and scrapers that interact with the platform. Airbnb labels its own auto-replies as coming from Airbnb, and its AI reply suggestions are visible only to the host.

**Does the EU AI Act make me tell guests they are talking to AI?**

Article 50(1) requires that people interacting with an AI system are informed of it, and Article 50(5) says clearly, at the latest at the first interaction. Formally the duty sits with the tool's provider, but the notice reaches the guest in your thread. When AI only drafts and you read and send, the Commission's guidelines treat it as indirect interaction and no notice is needed.

**From when does the AI Act disclosure duty apply?**

Article 50 applies from 2 August 2026 across the EU, per Article 113 and the European Commission's own Q&A. The Commission published its guidelines on Article 50 on 20 July 2026. Non-compliance can be fined up to 15 million euro or 3 percent of worldwide annual turnover, whichever is higher, with the lower of the two applying to small and medium-sized businesses.

**Does Booking.com require partners to label AI replies?**

We could not find such a requirement in the Partner Hub, checked on 26 September 2026. Booking.com offers „Help me reply”, which suggests a reply for you to review and send, and automatic replies, and says over 30 percent of partner responses are already automatic. What the Partner Hub does say is to communicate with guests only through Booking.com's platforms, the Extranet and Pulse.

**Do US hosts have to disclose AI in guest messages?**

It depends on the state. California's bot law makes it unlawful to use a bot to mislead a person in California about its artificial identity in order to incentivise a purchase, with a safe harbour for clear and conspicuous disclosure. Utah, after its 2025 amendment, requires disclosure when a consumer clearly and unambiguously asks, according to a Davis Polk summary. A one-line notice at the start covers both.


## Sources

- [EU AI Act, Article 50 (transparency obligations), text of Regulation 2024/1689 in the AI Act Explorer](https://artificialintelligenceact.eu/article/50/) (2026-09-26)
- [EU AI Act, Article 3 (definitions of provider and deployer)](https://artificialintelligenceact.eu/article/3/) (2026-09-26)
- [EU AI Act, Article 99 (penalties, paragraphs 4 and 6)](https://artificialintelligenceact.eu/article/99/) (2026-09-26)
- [EU AI Act, Article 113 (application from 2 August 2026)](https://artificialintelligenceact.eu/article/113/) (2026-09-26)
- [European Commission, Guidelines on the transparency obligations under Article 50 AI Act, C(2026) 5054, annex (PDF)](https://ec.europa.eu/newsroom/dae/redirection/document/131215) (2026-07-20)
- [European Commission, Q&A on transparency obligations under Article 50 of the AI Act](https://digital-strategy.ec.europa.eu/en/faqs/transparency-obligations-under-article-50-ai-act) (2026-09-26)
- [European Commission, Quick facts, transparency rules for AI systems](https://digital-strategy.ec.europa.eu/en/factpages/quick-facts-transparency-rules-ai-systems) (2026-09-26)
- [Airbnb, Content Policy (content generated or edited using AI)](https://www.airbnb.com/help/article/546) (2026-09-26)
- [Airbnb, Terms of Service (no bots or scrapers, section 11.1; Airbnb's own tools, section 9)](https://www.airbnb.com/help/article/2908) (2026-02-05)
- [Airbnb, About the updates to our Terms (no AI-generated evidence, effective dates)](https://www.airbnb.com/help/article/2877) (2026-09-26)
- [Airbnb, AI-powered features on Airbnb](https://www.airbnb.com/help/article/4097) (2026-09-26)
- [Airbnb Resource Center, Respond faster in the Messages tab (auto-replies identified as coming from Airbnb)](https://www.airbnb.com/resources/hosting-homes/a/respond-faster-in-the-messages-tab-770) (2026-05-20)
- [Airbnb Resource Center, Communicate more efficiently with messaging upgrades (AI suggestions visible only to the host)](https://www.airbnb.com/resources/hosting-homes/a/communicate-more-efficiently-with-messaging-upgrades-696) (2025-07-30)
- [404 Media, Airbnb hosts don't want to talk to guests anymore, are outsourcing messages to AI](https://www.404media.co/airbnb-hosts-dont-want-to-talk-to-guests-anymore-are-outsourcing-messages-to-ai/) (2026-04-14)
- [Hostrexa, Can AI respond to Airbnb guests? (secondary source)](https://www.hostrexa.com/blog/can-ai-respond-to-airbnb-guests) (2026-03-20)
- [HostReply, blog index (checked for a piece on AI disclosure rules)](https://hostreply.ai/blog) (2026-09-26)
- [Booking.com Partner Hub, Contacting guests („Help me reply”, Extranet and Pulse)](https://partner.booking.com/en-us/help/reservations/contact-extranet/contacting-guests) (2026-09-26)
- [Booking.com Partner Hub, setting up templates, automatic replies and message notifications](https://partner.booking.com/en-us/help/reservations/contact-extranet/setting-templates-automatic-replies-and-message-notifications) (2026-09-26)
- [Booking.com Partner Hub, Messaging tools (over 30 percent of responses are automatic replies)](https://partner.booking.com/en-gb/solutions/messaging-tools) (2026-09-26)
- [Booking.com newsroom, Smart Messenger and Auto-Reply](https://news.booking.com/bookingcom-debuts-agentic-ai-innovations-adding-to-its-robust-suite-of-genai-tools-for-customers/) (2025-10-09)
- [California Business and Professions Code, sections 17940 to 17943 (bot disclosure)](https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?lawCode=BPC&division=7.&title=&part=3.&chapter=6.&article=) (2026-09-26)
- [Davis Polk, Utah scales back reach of generative AI consumer protection law (SB 226, secondary source)](https://www.davispolk.com/insights/client-update/utah-scales-back-reach-generative-ai-consumer-protection-law) (2025-04-04)