Did the FCC Really Ban AI-Generated Voice Calls? No . . . But Restrictions Apply


3 minute read | February.09.2024

The Federal Communications Commission (FCC) has released a declaratory ruling confirming that calls containing AI-generated voices are subject to consent requirements imposed on “artificial or prerecorded voices” under the TCPA.

Although the headline of its press release proclaimed “FCC Makes AI-Generated Voices in Robocalls Illegal,” the FCC did not actually ban the use of the technology altogether.

What Is the FCC Trying to Accomplish? 

The FCC said the primary purpose of the declaratory ruling is to empower the FCC and state attorneys general to go after bad actors who use AI-generated voices in unsolicited robocalls to extort vulnerable family members, imitate celebrities and misinform voters.

The FCC has seen a rise in these types of unwanted calls over the last few years due in large part to advancements in AI-generated voices and increased availability of the technology. The FCC’s ruling aims to remove any ambiguity that calls containing AI-generated voices are outside the scope of the TCPA. To the contrary, the TCPA already addresses these types of calls through  obligations and restrictions on “artificial or prerecorded voices.”

What Does the New AI-Generated Voice Ruling Actually Mean?

The FCC confirmed that the TCPA’s restrictions on “artificial or prerecorded voices” encompass AI technologies that  simulate an artificial voice or resemble the voice of a real person (also known as “voice cloning” or an “audio deepfake”).

As a result, calls that use AI-generated voices are subject to a number of TCPA obligations and restrictions, including:

1. AI-Generated Voice Calls Require Consent.

The TCPA generally requires a caller to obtain:

  • Prior express consent before using an “artificial or prerecorded voice” in a non-marketing call to a cellular line.
  • Prior express written consent (a higher threshold than prior express consent) before using an “artificial or prerecorded voice” in marketing calls to residential or cellular lines.

Since calls using AI-generated voices are considered calls using an “artificial or prerecorded voice,” the TCPA requires some form of consent before placing such a call to an individual. Consent requirements under the TCPA are complex and nuanced; careful consideration of your specific scenario is required to determine the level of consent needed for your calls. 

2. AI-Generated Voice Calls Must Convey Certain Information and Options.

The TCPA imposes a number of other requirements for calls containing an “artificial or prerecorded voice,” including:

  • A clear statement of the identity of the business initiating the call.
  • A clear statement of the business’s phone number.
  • An automated, interactive voice- or key-activated opt-out mechanism allowing the person to make a do-not-call request (if delivering a marketing message or otherwise placing the call pursuant to certain consent exceptions).
  • An automated, interactive opt-out mechanism and instructions on how to opt-out of such calls in the future.

The TCPA is one of a host of federal and state telecommunications laws that impose obligations and restrictions on calls and other phone-based communications. Failing to comply can lead to expensive and time-consuming class action litigation.

Due to the complexity and litigious nature of the laws that may apply to AI-generated voice calls, we strongly recommend confirming compliance efforts with legal counsel. If you have questions about this development, reach out to the authors (Aravind Swaminathan, Shannon Yavorsky, Nick Farnsworth and Naomi Pen) or other members of the Orrick team.