Do Chatbots Have to Identify Themselves as Bots?
Educational information, not legal advice. Facts checked against the official law texts, current as of July 2026.
In California, the B.O.T. Act requires clear disclosure when a bot communicates with someone in the state with intent to mislead about its artificial identity, for the purpose of selling something or influencing a vote. The statute provides that a person using a bot is not liable under this section if the person discloses that it is a bot, clearly and conspicuously. This requirement applies to anyone communicating with people in California through a bot, no matter where the business is based.
Who the California B.O.T. Act Covers
The B.O.T. Act applies when someone uses a bot or AI chat agent to talk with people in California AND both of these are true: the person intends the recipient to be misled about whether the account is automated, and the purpose is to sell something or influence a purchase or a vote. A bot is defined as an automated online account where substantially all the actions are not coming from a real person.
The law does not target bots used for customer service, information, entertainment, or routine automation. It specifically addresses the scenario where a bot tries to persuade or sell while the recipient believes they are talking to a human.
The B.O.T. Act applies to any bot communicating with Californians that meets both the intent-to-mislead and commercial-or-electoral-purpose elements, whether on a website, in text messages, on social media, or in email.
What Clear and Conspicuous Disclosure Means
Clear and conspicuous means the person needs to know it is a bot without clicking links, scrolling, or reading fine print. The disclosure must be visible and understandable in the moment they are interacting with the bot. In a chat conversation, this typically means stating at the start or early in the exchange that they are talking to a bot.
Good disclosure examples include: "You are chatting with an automated bot," "This is a chatbot and not a person," or "Automated system: I am a bot." The language does not have to be fancy. It just needs to be plain English that a typical person understands immediately.
The law requires disclosure to be reasonably designed to inform the person they are talking to a bot. That means the disclosure needs to be in a place and format where the person will actually see it before they continue the conversation.
The Safe Harbor: Disclose and Liability Is Eliminated
California Business and Professions Code Section 17941 contains an explicit safe harbor. The statute provides that a person using a bot is not liable under this section if the person discloses that it is a bot, with disclosure clear and conspicuous and reasonably designed to inform.
This safe harbor does not require getting permission first or jumping through extra hoops. Disclosing the bot status in a clear, visible way meets the legal standard. The statute does not require proving the person understood or agreed; it requires that the disclosure was clear and reasonably designed to inform.
The safe harbor eliminates liability under the California unfair competition standard. It does not protect against other laws (like FTC rules on material connections or deepfakes), but it addresses the primary California B.O.T. Act risk.
Why Disclosure Matters Beyond California
Even if customers are not in California, disclosure aligns with broader regulatory trends. The FTC Endorsement Guides require clear disclosure when a sponsored post or endorsement includes substantive AI-generated content (not routine tool use). That rule applies nationwide. If a bot is writing sales pitches or reviews as part of a paid endorsement or material connection, the FTC expects disclosure of both the material relationship and the AI involvement.
The EU AI Act Article 50 takes effect August 2, 2026. It requires disclosure when AI generates or manipulates text, images, audio, or video published online. If anyone in the EU can see content, compliance is required. The trend across multiple jurisdictions is clear: disclosure of AI involvement is becoming a baseline requirement.
Disclosure also builds trust. People are more likely to engage with a bot they know is a bot than one they discover is fake after committing time or money. Transparent disclosure reduces friction and improves user relationships.
Upcoming Changes and Multi-Jurisdiction Compliance
California SB 942, the AI Transparency Act, takes effect August 2, 2026. SB 942 targets companies that build publicly accessible generative AI systems with over one million monthly users. It does not apply to small businesses or creators merely using AI tools. If a business uses ChatGPT or Claude to write messages, SB 942 does not apply to that business. It applies to OpenAI, Anthropic, and other large AI system providers.
The trend across California, the EU, and US federal enforcement is toward mandatory disclosure of AI involvement. The B.O.T. Act has been in force since 2019. The new rules reinforce the pattern: transparency is becoming the legal standard. Businesses using bots to communicate with people about products or votes should disclose the automated nature. That approach keeps operations aligned with current requirements and ready for future rules.
Frequently asked questions
Does the B.O.T. Act apply to chatbots used only for customer service?
The B.O.T. Act specifically targets bots used to sell something or influence a vote with intent to mislead about the bot's identity. Bots answering FAQs, helping with returns, or providing information without a sales or electoral purpose are not the focus of the statute. However, other laws (like FTC rules on endorsed content or the EU AI Act) may apply depending on what the bot does and where its users are located.
What if someone does not read the disclosure?
Under the B.O.T. Act safe harbor, if disclosure was clear, conspicuous, and reasonably designed to inform, the legal standard is met. Liability is eliminated by following the statute's safe harbor provision. The law does not require verifying that the person actually read the disclosure, only that it was visible and understandable to a typical person in that context.
Do I need to disclose if I am using AI to write content that a human then sends?
That depends on the specifics. If a person wrote the message and sent it through an automated system, the B.O.T. Act may not apply if the bot is not substantially automating the actions and there is no intent to mislead about the artificial identity. However, FTC rules on material connections in endorsements and the EU AI Act rules on AI-generated content may apply separately. The safest practice is to consider disclosure when a bot is communicating in a commercial or electoral context.
What is the difference between the B.O.T. Act and California SB 942?
The B.O.T. Act (in force since 2019) applies when someone uses a bot with intent to mislead about its artificial identity for sales or votes, and it targets anyone using such a bot. SB 942 (taking effect August 2, 2026) targets the companies that build large generative AI systems with over one million monthly users. It requires those providers to offer detection tools and labeling. If a business uses a third-party chatbot platform, the platform handles SB 942 compliance. The business handles B.O.T. Act disclosure.
Which of these laws apply to you?
The free checker asks eight yes-or-no questions and maps your answers to all five AI disclosure laws, with sample wording for anything that needs action.
Run the free checkThis page is educational information, not legal advice. Laws change and your situation may differ. See the law guides for official sources, and consult a licensed attorney for advice about your specific situation.