Overview
The Federal Communications Commission (FCC) recently declared that AI-generated voices do fall within the meaning of “artificial” under the Telephone Consumer Protection Act (TCPA) in a ruling issued on February 8, 2024.1 The ruling follows a recent surge in robocalling and robotexting using AI technologies that simulate a human voice or that generate call content using a prerecorded voice, resulting in a significant number of complaints from consumers and attorneys general. Going forward, businesses and callers who use AI-generated voices will need to ensure they first obtain, at a minimum, prior express consent from the receiving party and otherwise comply with the requirements of the TCPA.
Background
The TCPA generally prohibits telephone calls made to residential and mobile phone lines without some form of consent if the calls are placed using an “artificial or pre-recorded voice.”2 The FCC has confirmed that “calls” include text messages.3 The TCPA does not define the terms “artificial” or “prerecorded voice.”
On November 16, 2023, the FCC issued a Notice of Inquiry (NOI) to gather public comment on the implications of emerging AI technologies as part of an ongoing effort to protect consumers from unwanted and illegal telephone calls and text messages under the TCPA.4 The FCC’s NOI requested input on a broad range of issues, such as defining AI in the context of robocalls and robotexts, the current state of AI use in this context, the impact of emerging AI technologies on consumer privacy rights under the TCPA, the FCC’s next steps to address these issues, and whether AI technology can minimize the nuisances associated with the use of artificial or prerecorded voice messages by acting as the functional equivalent to calls with live agents.5
Recent Rise in Robocall and Robotexts That Use AI Technologies
In its NOI, the FCC noted that robocalls and robotexts are consistently the top category of consumer complaints it receives.6 In particular, the NOI was released following a robocall campaign that used a deepfake of President Biden’s voice to tell voters in New Hampshire to stay home from the January Republican primary election.7 There have also been numerous incidents of bad actors using AI to imitate the voice of a loved one to convince the call recipient to do something they normally would not do. An example of this is using the voice of a grandchild to make a grandparent believe the grandchild is on the phone begging for funds when in reality it is a bad actor preying on the grandparent’s emotions.8
“AI-generated voice cloning and images are already sowing confusion by tricking consumers
into thinking scams and frauds are legitimate,” said FCC Chairwoman Jessica Rosenworcel. “No matter what celebrity or politician you favor, or what your relationship is with your kin when they call for help, it is possible we could all be a target of these faked calls.”9
In response to the NOI, several commenters, including 26 attorneys general, asked the FCC to confirm that AI technologies that simulate a human voice or generate call content using a prerecorded voice should be considered “artificial or prerecorded voices” for purposes of the TCPA.10
The FCC’s Declaratory Ruling
In its Declaratory Ruling released last week, the FCC has now confirmed that current AI technologies, including “voice cloning,” that resemble human voices or generate call content using a prerecorded voice fall within the meaning of “artificial or prerecorded voice” under the TCPA.11 The FCC explained that voice cloning and other similar technologies are “artificial” voice messages because a person is not speaking, and they are therefore the types of calls from which the TCPA seeks to protect consumers.12
The FCC’s Declaratory Ruling makes clear that the “TCPA does not allow for any carve out of technologies that purport to provide the equivalent of a live agent, thus preventing unscrupulous businesses from attempting to exploit any perceived ambiguity in [the] TCPA rules.”13
As a result of the FCC’s Declaratory Ruling clarifying that the TCPA’s restrictions apply to calls and prerecorded messages using AI technologies, state attorneys general across the country can now seek damages from callers who violate the law by misusing these types of calls.14
Implications
The implications of the FCC’s Declaratory Ruling are significant for businesses using AI-generated voices in marketing and promotional efforts, or to communicate with consumers even for informational purposes only. Because these technologies now clearly fall within the purview of the TCPA, companies should be aware that they may be required to obtain at least prior express consent from the recipient, absent an emergency purpose or statutory exemption, as well as to comply with any other relevant TCPA requirements based on the substance of the call.15
Additionally, as with any calls or messages that use artificial or prerecorded voices, when using AI-generated voices, businesses must disclose their identify at the beginning of any such call or message, and clearly state their telephone number during the call or message.16 When these types of calls or messages include an advertisement, promotion, or otherwise constitute telemarketing, they also must provide an opt-out mechanism for the called or texted person to make a do-not-call request within two seconds of providing the required disclosures.17
Failure to comply with the requirements of the TCPA can leave companies vulnerable to distracting and expensive litigation by consumers, the FCC, and state attorneys general, as well as high statutory fines. We highly recommend that companies using AI-generated voices in their business communications review their current practices to ensure compliance these new developments.
[1] Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Declaratory Ruling, FCC 24-17 (rel. Feb. 8, 2024).
[2] 47 U.S.C. § 227(b)(1).
[3] See Rules and Regulations Implementing the Telephone Consumer Protection Act of 1991, CG Docket No. 02-278, Report and Order, 18 FCC Rcd 14014 (2003).
[4] Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Notice of Inquiry, FCC 23-101 (rel. Nov. 16, 2023) (AI NOI).
[5] Id.
[6] Id.
[7] See Jared Foretek, FCC Targeting Companies Behind Biden Deepfake In NH, Law360 (Feb. 6, 2024), https://www.law360.com/articles/1794738/fcc-targeting-companies-behind-biden-deepfake-in-nh.
[8] Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Statement of Chairwoman Jessica Rosenworcel, FCC 24-17 (rel. Feb. 8, 2024).
[9] FCC Chairwoman: Make AI Voice-Generated Robocalls Illegal, FCC News (Jan. 31, 2024), https://docs.fcc.gov/public/attachments/DOC-400212A1.pdf.
[10] Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Notice of Inquiry, FCC 23-101 (rel. Nov. 16, 2023) (AI NOI); Reply Comments of 26 State Attorney Generals, available at https://www.fcc.gov/ecfs/document/10116069739204/1.
[11] Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Declaratory Ruling, FCC 24-17 (rel. Feb. 8, 2024).
[12] Id.
[13] Id.
[14] Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Statement of Chairwoman Jessica Rosenworcel, FCC 24-17 (rel. Feb. 8, 2024).
[15] See 47 CFR § 64.1200(a)(1), (3).
[16] See 47 CFR § 64.1200(b)(1), (2).
[17] See 47 CFR § 64.1200(b)(3).
