Best CRM AI
Home / AI disclosure rules for CRM teams

AI disclosure rules for CRM teams

Answer in brief

As of 8 November 2026, this table lists 9 rules on disclosing AI in voice, chat and email, each read at the statute, regulation or regulator page, across United States federal rules, the European Union, Maine, Utah and Washington.

Content reviewed 8 November 2026. How this is assessed.

Fennick and this reference are run by the same operator. Fennick is not assessed on this page. Fennick has no published client list.

The rules table

Each row was read at the statute, regulation or regulator page on the date shown. Law firm and vendor posts were used only to find the source. Where a page did not state something, the row says so.

9 rules read on 8 November 2026. Not legal advice.

Jurisdiction and ruleChannelTriggerDutyEffective date and amendmentsCitationDate readWhat a CRM admin should switch on or check
United States (federal), FCCVoiceAn outbound call that uses an artificial or prerecorded voice, including a voice generated by AI, unless an emergency purpose or exemption applies.Get the called party's prior express consent. Written consent where the call advertises or telemarkets. Identify the entity responsible for the call. Offer the specified opt-out on advertising or telemarketing calls.Effective on release, 8 Feb 2024 (adopted 2 Feb 2024).FCC 24-17, paras 2, 5, 9, note 138 Nov 2026Turn on a consent field per phone number with type, date and source. Block any AI voice campaign to numbers without it. Put the caller identification and opt-out in the voice script.
United States (federal), FTC Telemarketing Sales RuleVoiceA telemarketing call that delivers a prerecorded message to sell goods or services. The FTC guide read does not mention AI and does not say how the rule treats a synthetic voice.Hold a prior signed written agreement from that seller. State the seller's identity, the sales purpose and the nature of the goods. Offer an automated opt-out right after those statements. Start the message within two seconds of the greeting.Guide says the amendments were adopted Aug 2008. It gives no effective date for them.FTC guide, prerecorded messages section8 Nov 2026Store the signed agreement against the number. Do not accept agreements obtained by a lead seller. Map opt-out events to your do-not-call list.
United States (federal), FTC CAN-SPAM guideEmailAny email whose primary purpose is the commercial advertisement or promotion of a commercial product or service. The FTC guide read does not mention AI.Truthful headers and subject lines, identify the message as an ad, include a valid postal address, tell recipients how to opt out, honor opt-outs within 10 business days. The guide says you cannot contract away the responsibility.Guide dated Aug 2023, penalty figures edited Jan 2024.FTC CAN-SPAM compliance guide8 Nov 2026Lock the sender, address and opt-out footer so an AI draft cannot edit them. See AI-drafted email and the CAN-SPAM rules.
Maine, 10 M.R.S. section 1500-DDChat, voice. The rule also reaches "any other computer technology".Using a chatbot or other computer technology in trade and commerce with a consumer in a way that may mislead a reasonable consumer into believing they are engaging with a human.Notify the consumer in a clear and conspicuous manner that they are not engaging with a human being. A violation is a violation of the Maine Unfair Trade Practices Act.Signed 12 Jun 2025. Maine's general effective date for nonemergency First Special Session laws is 24 Sep 2025, and the bill page does not mark this law an emergency.PL 2025 c. 294, 10 M.R.S. section 1500-DD8 Nov 2026Show a clear not-human notice in every chat greeting and voice opening for Maine consumers, and record that it was shown.
Utah, Utah Code 13-77-103(1)Chat, voice, visualA supplier uses generative AI to interact with an individual in connection with a consumer transaction, and the individual makes a clear and unambiguous request to find out whether it is AI or a human.Disclose that the individual is interacting with generative AI and not a human.Effective 7 May 2025. Enacted by Chapter 465, 2025. Definitions in 13-77-101 were amended by Chapter 95, 2026.Utah Code 13-77-1038 Nov 2026Make sure the assistant answers truthfully when asked if it is AI. Section 13-77-104 gives a safe harbor if the AI discloses at the outset and throughout, so check that a banner persists.
Utah, Utah Code 13-77-103(2) and (3)Chat, voiceAn individual in a regulated occupation provides services through generative AI and the use is a high-risk AI interaction, such as collecting financial, health or biometric data or giving advice relied on for significant decisions.Prominently disclose the use: verbally at the start of a verbal interaction, in writing before the start of a written one.Effective 7 May 2025. Enacted by Chapter 465, 2025.Utah Code 13-77-1038 Nov 2026If licensed staff use the CRM assistant with clients, mark those workflows and require the opening disclosure.
European Union, AI Act Article 50(1) and (5)Chat, voiceAn AI system intended to interact directly with natural persons, unless it is obvious to a reasonably observant person.The provider must design the system so people know they are dealing with AI, at the latest at the time of first interaction.Applies from 2 Aug 2026 (Recital 179). Regulation (EU) 2026/1744 replaced Article 50(7) and left this paragraph unchanged in the text read.Regulation (EU) 2024/1689, Article 508 Nov 2026Ask the provider in writing how it meets Article 50(1). Check the first message of each EU-facing chat or voice flow names the system as AI.
European Union, AI Act Article 50(2)Email, chat, voice contentA provider's AI system generates synthetic audio, image, video or text. Assistive editing tools and systems that do not substantially alter the input are excepted.Mark outputs in a machine-readable format so they are detectable as artificially generated or manipulated.Applies from 2 Aug 2026 (Recital 179). Systems placed on the market before 2 Aug 2026 must comply by 2 Dec 2026, under Article 111(4) added by Regulation (EU) 2026/1744.Regulation (EU) 2024/1689, Article 508 Nov 2026Ask the provider whether generated email and call audio carry a marking, and whether your templates or export path strip it.
Washington, ESHB 2225 (Chapter 168, Laws of 2026)Chat and other natural language interfacesAn AI companion chatbot made available to users in Washington. A bot used only for customer service or business operations is excluded if it does not sustain a relationship across interactions and generate outputs likely to elicit emotional responses.Clear and conspicuous disclosure that the chatbot is artificially generated and not human, at the start and at least every three hours. Reasonable measures to stop it claiming to be human.Takes effect 1 Jan 2027.ESHB 2225, sections 2 and 38 Nov 2026Record why your assistant falls in the exclusion. Check memory across conversations and any relationship-style persona, since those bear on it.

Do you have to disclose AI voice calls?

In the United States, the FCC's February 2024 ruling treats AI-generated voices as artificial voices under the Telephone Consumer Protection Act. The duties the ruling records are prior express consent, identification of the entity responsible and, on advertising or telemarketing calls, specified opt-outs. The ruling does not add a separate duty to announce that the voice is AI.

State rules can add that duty. Maine's rule covers aural communication and requires a clear notice where a reasonable consumer could be misled. Utah's rule covers audio and is triggered when the person clearly asks. The EU's Article 50(1) covers systems that interact directly with people.

Do chatbots have to identify themselves?

It depends on where the consumer is and what the bot does. In Maine, yes, wherever a reasonable consumer could be misled into believing they are engaging with a human. In Utah, a supplier must answer truthfully when clearly asked, and a safe harbor applies if the AI discloses at the outset and throughout. In the EU, providers must ensure people know unless it is obvious.

Washington's 2026 law is aimed at AI companion chatbots and excludes a bot used only for customer service or business operations, so long as it does not sustain a relationship across interactions and generate outputs likely to elicit emotional responses.

Do AI-written emails have to say they were written by AI?

The two FTC guides read do not mention AI. The CAN-SPAM guide applies to commercial email whoever or whatever drafted it, so an AI draft carries the same header, address and opt-out duties. Article 50(2) of the EU AI Act is different: it requires providers to mark generated text in a machine-readable format, which is a duty on the provider rather than a label in the message.

Each rule in one sentence

United States, FCC: under the FCC's February 2024 ruling (FCC 24-17), an outbound call using an AI-generated voice is a call using an artificial voice, so it needs the called party's prior express consent and must identify the entity responsible.

United States, FTC: the FTC's Telemarketing Sales Rule guide requires a prior signed written agreement, seller identification and an automated opt-out for prerecorded telemarketing calls, and the guide does not address AI.

United States, FTC: the FTC's CAN-SPAM guide requires truthful headers, an ad identification, a postal address and opt-outs honored within 10 business days for commercial email, however the email was drafted, and the guide does not address AI.

Maine: 10 M.R.S. section 1500-DD bars using a chatbot or other computer technology in trade and commerce in a way that may lead a reasonable consumer to believe they are engaging with a human, unless the consumer is clearly and conspicuously notified.

Utah: Utah Code 13-77-103(1) requires a supplier using generative AI in a consumer transaction to disclose that it is AI when the individual clearly and unambiguously asks.

Utah: Utah Code 13-77-103(2) and (3) require individuals in regulated occupations to prominently disclose generative AI in high-risk interactions, verbally at the start of a verbal interaction and in writing before a written one.

European Union: Article 50(1) of the AI Act requires providers to ensure people know they are dealing with an AI system, unless that is obvious, with the rule applying from 2 August 2026.

European Union: Article 50(2) of the AI Act requires providers of systems generating synthetic text, audio, image or video to mark outputs in a machine-readable format, with systems already on the market before 2 August 2026 given until 2 December 2026.

Washington: ESHB 2225 requires operators of AI companion chatbots to disclose that the chatbot is not human at the start and at least every three hours from 1 January 2027, and it excludes customer service bots that do not sustain a relationship and elicit emotional responses.

Which jurisdictions were reviewed?

Reviewed and listed here: United States federal rules from the FCC and FTC, the European Union, Maine, Utah and Washington.

The official text could not be opened on 8 November 2026 for California, Colorado, Oregon, Idaho, Nebraska or Hawaii. For Iowa, the enrolled text of Senate File 2417 was read, but the copy showed a blank approval line and the governor's action could not be confirmed, so it is not listed. Nothing is said here about those jurisdictions.

No other jurisdiction was reviewed. A jurisdiction missing from this table has not been reviewed. Its absence does not mean it has no rule.

What this page does not do

This page is general information, not legal advice. It does not say whether any rule applies to your business. Ask counsel where you are unsure.

The current federal rule text at 47 CFR 64.1200 could not be opened on the date read, so the FCC row relies on the ruling's description of those rules. Check the current regulation before relying on it.

Change log

8 November 2026: page published with 9 rows, all read on that date. Recorded on publication: Regulation (EU) 2026/1744 amending the AI Act, Utah's 2026 amendment to Chapter 77 definitions, and the Maine reallocation of this section from 10 M.R.S. section 1500-Y.

Sources for this entry

Continue this reading path

Cite this entry

Best CRM AI. "AI disclosure rules for CRM teams." Content reviewed 8 November 2026. https://bestcrmai.com/ai-disclosure-rules/

General information only. This is not legal advice. Rules and vendor terms change. Confirm current terms with the provider and take advice on your own obligations.