What should an AI call disclose before the conversation starts

Tell the person that the call is automated, identify the business and explain the purpose in plain language. Make opt out and human help easy to use. Before placing a call, confirm the consent, calling and recording rules for the recipient and use case because disclosure alone does not make a call lawful.

What should an AI call disclose before the conversation starts

Explain the automated nature early

Give the disclosure before the system asks for information or starts a sales message. Use ordinary words such as This is an automated call from DripTell rather than technical model names. If the voice imitates a real person, the risk of confusion is higher and requires specialist legal review.

  • Say that the caller is automated
  • Name the business responsible for the call
  • Explain the purpose before requesting information

Check permission before placing the call

Determine where the recipient is located, whether the call is sales or service, how the number was obtained and what consent evidence exists. Also check calling hours, do not call requests and any sector rules. A generic disclosure script cannot replace this review.

  • Keep the source, time and purpose of consent
  • Apply the current opt out before every outbound call
  • Separate service calls from sales and upsell activity

Make opt out and human help work during the call

Tell the person how to stop the automated call and use simple spoken choices or keypad input where required. Confirm the request, end the automated flow and prevent another call when the person opts out. Offer a human route when the system is uncertain or the person asks for one.

  • Recognize clear spoken requests to stop
  • Apply the preference across connected campaigns
  • Transfer the context instead of making the person repeat it

Keep evidence for every call version

Store the script version, voice configuration, consent record, call time, disclosure completion, opt out result and handoff outcome. Review failed disclosures and calls that continued after an opt out. Evidence should help a reviewer reconstruct what the person heard and what the system did next.

  • Version the disclosure and call purpose
  • Record the source decision without storing unnecessary call content
  • Audit opt out and transfer failures

Review the rules for the target market

In the United States, FCC guidance treats AI generated voices as artificial or prerecorded voice under the Telephone Consumer Protection Act. FTC telemarketing rules can also apply to sales calls and prerecorded messages. Other countries and sectors use different consent, identity and recording rules. Get qualified legal advice for the actual campaign.

  • Check the current regulator and telemarketing rules
  • Review local recording and privacy requirements
  • Repeat the review when the purpose, market or voice changes

Primary references https://docs.fcc.gov/public/attachments/FCC-24-17A1_Rcd.pdf and https://www.ftc.gov/business-guidance/resources/complying-telemarketing-sales-rule

Questions teams ask before they connect the workflow.

Is saying that a call uses AI enough for compliance?

No. The business must also check consent, calling purpose, opt out, recording, identity and market specific rules. Disclosure is one control, not complete compliance.

When should an AI call offer a human agent?

Offer a human route when the person asks, the system is uncertain, the request is sensitive or the next action requires judgment or stronger verification.

What evidence should be kept for an automated call?

Keep the consent source, call purpose, script version, call time, disclosure result, opt out result and any transfer outcome for the period required by the applicable policy and law.

Design the disclosure and human route before calling

Bring the target market, purpose, consent source and exception cases. DripTell can help map the operational controls for specialist review.