Federal Communications Commission	FCC 24-17


Before the
FEDERAL COMMUNICATIONS COMMISSION
WASHINGTON, D.C. 20554


In the Matter of

Implications of Artificial Intelligence Technologies
on Protecting Consumers from Unwanted Robocalls and Robotexts
)
)
)
)
)


CG Docket No. 23-362


DECLARATORY RULING

Adopted:  February 2, 2024	Released:  February 8, 2024

By the Commission:  Chairwoman Rosenworcel and Commissioners Starks and Gomez issuing separate statements.

I. INTRODUCTION
1. As artificial intelligence (AI) technologies emerge that affect the existing regulatory landscape that protects consumers from unwanted and illegal robocalls, we take action today to ensure that consumers continue to receive the protections afforded under the Telephone Consumer Protection Act (TCPA). See Telephone Consumer Protection Act of 1991, Pub. L. No. 102-243, 105 Stat. 2394 (1991), codified at 47 U.S.C. § 227.  As discussed in more detail below, the TCPA regulates any call made using an “automatic telephone dialing system” or an artificial or prerecorded voice.  Any such call is considered a “robocall” for purposes of this proceeding.  See 47 U.S.C. § 227(b)(1).  
  AI technologies that can generate content such as artificial voice communications can be beneficial, but also pose new challenges to consumers. See, e.g., Executive Order No. 14110, Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, 88 Fed. Reg. 75191 (Oct. 30, 2023).
  We take this opportunity to clarify the application of the TCPA to these technologies.
2. In this Declaratory Ruling, we confirm that the TCPA’s restrictions on the use of “artificial or prerecorded voice” encompass current AI technologies that generate human voices. See 47 U.S.C. § 227(b); 47 CFR § 64.1200(a)(1), (3).
  As a result, calls that use such technologies fall under the TCPA and the Commission’s implementing rules, and therefore require the prior express consent of the called party to initiate such calls absent an emergency purpose or exemption. The Commission has exercised its statutory authority to adopt certain exemptions from this prohibition.  See 47 CFR § 64.1200(a)(3), (9). 

II. BACKGROUND
3. The TCPA protects consumers from unwanted calls made using an artificial or prerecorded voice. See 47 U.S.C. § 227(b)(1).
  In relevant part, the TCPA prohibits initiating “any telephone call to any residential telephone line using an artificial or prerecorded voice to deliver a message without the prior express consent of the called party” unless a statutory exception applies or the call is “exempted by rule or order by the Commission under [section 227(b)(2)(B)].” 47 U.S.C. § 227(b)(1)(B).  The TCPA does not define the terms “artificial” or “prerecorded voice.” 
  The TCPA also prohibits, without the prior express consent of the called party, any non-emergency call made using an automatic telephone dialing system or an artificial or prerecorded voice to certain specified categories of telephone numbers including emergency lines and wireless numbers. See 47 U.S.C. § 227(b)(1)(A).  
  
4. On November 16, 2023, the Commission released a Notice of Inquiry seeking to better understand the implications of emerging AI technologies as part of our ongoing efforts to protect consumers from unwanted and illegal telephone calls and text messages under the TCPA. See 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).
  As part of that inquiry, the Commission indicated that “the Commission’s authority under the TCPA encompasses current uses of AI in robocalling and robotexting as we understand them, e.g., emulating human speech and interacting with consumers as though they were live human callers when generating voice and text messages.” Id. at para. 9.  
  In addition, the Commission stated “that certain AI technologies such as ‘voice cloning’ appear to fall within the TCPA’s existing prohibition on artificial or prerecorded voice messages because this technology artificially simulates a human voice” and sought comment on these issues. Id. at para. 25.
  In response, several commenters ask the Commission to confirm that AI technologies that simulate a human voice and/or generate call content using a prerecorded voice should be considered an “artificial or prerecorded voice” for purposes of the TCPA. See, e.g., State AG Reply Comments at 2-3 (“any type of AI technology that generates a human voice should be considered an ‘artificial voice’ for purposes of the TCPA”); Transaction Network Services Comments at 4 (“The Commission should also take this opportunity to issue clear and explicit guidance to the industry explaining that the plain language of the TCPA applies to any AI-generated voice call”); USTelecom Comments at 3-4 (“The Commission, however, should take this opportunity to issue clear and explicit guidance to the industry confirming that the plain language of the [TCPA] applies to any AI-generated voice call”). 
 
III. DISCUSSION
5. Consistent with our statements in the AI NOI, we confirm that the TCPA’s restrictions on the use of “artificial or prerecorded voice” encompass current AI technologies that resemble human voices and/or generate call content using a prerecorded voice. See AI NOI at paras. 9, 12, 25.
  Therefore, callers must obtain prior express consent from the called party before making a call that utilizes artificial or prerecorded voice simulated or generated through AI technology. If these robocalls introduce an advertisement or contain telemarketing, the Commission’s rules require that the caller obtain the prior express written consent of the called party.  See 47 CFR § 64.1200(a)(2), (3).
  AI technologies such as “voice cloning” fall within the TCPA’s existing prohibition on artificial or prerecorded voice messages because this technology artificially simulates a human voice. See Anisha Kohli, From Scams to Music, AI Voice Cloning is on the Rise (Apr. 29, 2023) https://time.com/6275794/ai-voice-cloning-scams-music.
  Voice cloning and other similar technologies emulate real or artificially created human voices for telephone calls to consumers. AI NOI at para. 25.
  They are “artificial” voice messages because a person is not speaking them, This accords with court decisions that have described an “artificial voice” under the TCPA as including “a sound resembling a human voice that is originated by artificial intelligence.”  Trim v. Reward Zone USA, 76 F.4th 1157, 1163 (9th Cir. 2023).
 and, as a result, when used they represent the types of calls the TCPA seeks to protect consumers from. 47 U.S.C. § 227(b)(1)(A), (b)(1)(B) (imposing restrictions on calls “using . . . an artificial . . . voice”).
  The Commission also recognized in the AI NOI the potential for AI technologies to communicate with consumers using prerecorded voice messages. AI NOI at para. 13.
  Where this occurs, we conclude that the call is “using” a “prerecorded voice” within the meaning of the TCPA. 47 U.S.C. § 227(b)(1)(A), (b)(1)(B) (imposing restrictions on calls “using” a “prerecorded voice”).
  In reaching these conclusions, we agree with several commenters that the existing consumer protections under the TCPA apply to these calls. See, e.g., INCOMPAS Reply Comments at 4; State AG Reply Comments at 2-3; Transaction Network Services Comments at 4; USTelecom Comments at 3-4.
  No commenter opposes these conclusions.
6. Our finding will deter negative uses of AI and ensure that consumers are fully protected by the TCPA when they receive such calls.  And it also 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 our TCPA rules. See State AG Reply Comments at 2.
  Although voice cloning and other uses of AI on calls are still evolving, we have already seen their use in ways that can uniquely harm consumers and those whose voice is cloned. See, e.g., Ali Swenson and Will Weissert, New Hampshire investigating fake Biden robocall meant to discourage voters ahead of primary (Jan. 24, 2024) https://apnews.com/article/new-hampshire-primary-biden-ai-deepfake-robocall-f3469ceb6dd613079092287994663db5; Justin Green and Allie Weintraub, Experts warn of rise in scammers using AI to mimic voices of loved ones in distress (July 7, 2023). https://abcnews.go.com/Technology/experts-warn-rise-scammers-ai-mimic-voices-loved/story?id=100769857.
  Voice cloning can convince a called party that a trusted person, or someone they care about such as a family member, wants or needs them to take some action that they would not otherwise take.  Requiring consent for such calls arms consumers with the right not to receive such calls or, if they do, the knowledge that they should be cautious about them. As discussed below, section 64.1200(b)(1) of the Commission’s rules provides: “All artificial or prerecorded voice telephone messages shall: (1) At the beginning of the message, state clearly the identity of the business, individual, or other entity that is responsible for initiating the call. . . .”
  
7. To be clear, we understand that not all AI-generated calls may be deceptive or annoy all consumers.  But the TCPA’s demands fully apply to those calls and, thus, consumers can themselves choose whether to receive them.  Indeed, as Congress observed in adopting the TCPA, “residential telephone subscribers consider automated or prerecorded telephone calls, regardless of the content or the initiator of the message, to be a nuisance and an invasion of privacy.” Telephone Consumer Protection Act of 1991, Public Law 102-243, 102d Congress, § 2(10) (1991).
  Our ruling today thus comports with both the statutory text and the legislative intent of the TCPA and protects consumers from unwanted and illegal prerecorded or artificial voice calls.
8. Our conclusion is also consistent with the Commission’s Soundboard Ruling, which confirmed that the TCPA applies to any telephone call that is initiated using an artificial or prerecorded voice message. See Rules and Regulations Implementing the Telephone Protection Act of 1991, NorthStar Alarm Services LLC’s Petition for Expedited Declaratory Ruling, Yodel Technologies Petition for Expedited Declaratory Ruling or in the Alternative Retroactive Waiver, CG Docket No. 02-278, Declaratory Ruling and Order, 35 FCC Rcd 14640 at 14643, para. 11 (2020) (Soundboard Ruling).  Although certain provisions of the TCPA use the term “make” rather than “initiate” a call, see, e.g., 47 U.S.C. § 227(b)(1)(A), “[t]he Commission uses the words ‘make’ and ‘initiate’ interchangeably in its rules arising under § 227(b)(1)(A) without drawing any distinction.”  Dialing Services, LLC, Notice of Apparently Liability for Forfeiture, 29 FCC Rcd 5537, 5542, para 14, n.30 (2014); see also Rules and Regulations Implementing the Telephone Consumer Protection Act of 1991, et al., Declaratory Ruling and Order, 30 FCC Rcd 7961, 7979, para. 27 n.95 (2015) ( “the Commission has recently stated that the same logic that applies to the ‘initiation’ of calls under section 227(b)(1)(b) applies to the ‘making’ of calls under section 227(b)(1)(A)”).
  There the Commission confirmed that the presence of a live agent on a call selecting the prerecorded messages to be played “does not negate the clear statutory prohibition against initiating a call using a prerecorded or artificial voice.” Soundboard Ruling at 14644, para. 12.
  We find that this rationale applies to AI technologies, including those that either wholly simulate an artificial voice or resemble the voice of a real person taken from an audio clip to make it appear as though that person is speaking on the call to interact with consumers.  In both cases, the call is initiated using an artificial or prerecorded voice under the TCPA.
9. As a result, callers that use such technologies must obtain the prior express consent of the called party to initiate such calls absent an emergency purpose or exemption. See 47 CFR § 64.1200(a)(1), (3).  
  In addition, our rules require that all artificial or prerecorded voice messages must provide certain identification and disclosure information for the entity responsible for initiating the call. See 47 CFR § 64.1200(b)(1), (2).
  In every case where the artificial or prerecorded voice message includes or introduces an advertisement or constitutes telemarketing, it must also offer specified opt-out methods for the called party to make a request to stop calling that telephone number. See 47 CFR § 64.1200(b)(3).
  These requirements are applicable to any AI technology that initiates any outbound telephone call using an artificial or prerecorded voice to consumers.
IV. ORDERING CLAUSE
10. Accordingly, IT IS ORDERED, pursuant to Sections 1-4, and 227 of the Communications Act of 1934, as amended, 47 U.S.C. §§ 151-154, 227, sections 1.2 and 64.1200 of the Commission’s rules, 47 CFR §§ 1.2 and 64.1200, that this Declaratory Ruling IS ADOPTED.
11. IT IS FURTHER ORDERED that this Declaratory Ruling shall be effective upon release.

						FEDERAL COMMUNICATIONS COMMISSION




						Marlene H. Dortch
						Secretary

2
	Federal Communications Commission	FCC 24-17


STATEMENT OF
CHAIRWOMAN JESSICA ROSENWORCEL
Re:	Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Declaratory Ruling (February 2, 2024)
It seems like something from the far-off future, but it is already here.  
Artificial Intelligence-generated voice cloning and image creating tools are now more accessible and more likely to be used for fraud.  This technology can confuse us when we listen, view, and click, because it can trick us into thinking all kinds of fake stuff is legitimate.  Already we see this happening with Tom Hanks hawking dental plans online, a vile video featuring Taylor Swift, and calls from candidates for political office that are designed to confuse us about where and when to vote.  We also have stories of grandparents who are led to believe that it is really their grandchild on the phone begging for funds, only to learn later it was a bad actor preying on their willingness to forward money to family.
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 on the receiving end of these faked calls.  That is why at the Federal Communications Commission, we are taking action to stop the bad actors responsible for this junk.  In November we launched a Notice of Inquiry to build a record to help combat illegal robocalls where AI might be involved.  We asked questions about how AI might be used for scams that arise out of robocalls, by mimicking the voices of those we know.  But just as important, we asked about how AI can help us with pattern recognition so that we turn this technology into a force for good that can recognize illegal robocalls before they ever even reach us on the phone.
Today we go a step further.  We issue a Declaratory Ruling to make clear that under the Telephone Consumer Protection Act, restrictions on the use of an “artificial or prerecorded voice” apply to the AI technologies that are already being used to generate unwanted and unlawful robocalls.
What does that mean?  The Telephone Consumer Protection Act is the primary law we have to help limit unwanted robocalls.  It means that AI technologies like voice cloning fall within this law’s existing prohibitions and that calls that use this technology to simulate a human voice are illegal, unless callers have obtained prior express consent.  It also means that when these calls happen, State Attorneys General across the country can go after the bad actors behind these robocalls and seek damages under the law.  In fact, 26 State Attorneys General—from all across the country—wrote to us supporting this approach.  Even better, at the Commission we now have a Memorandum of Understanding with 48 State Attorneys General who have agreed to work with us to combat robocalls.  Now, with this Declaratory Ruling, we will have another tool to go after voice cloning scams and get this junk off the line.


2
	Federal Communications Commission	FCC 24-17


STATEMENT OF 
COMMISSIONER GEOFFREY STARKS 
Re:	Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Declaratory Ruling (February 2, 2024)
We all know unwanted robocalls are a scourge on our society.  But I am particularly troubled by recent harmful and deceptive uses of voice cloning in robocalls.  Real world examples here are no longer theoretical.  Bad actors are using voice cloning – a generative AI technology that uses a recording of a human voice to generate speech sounding like that voice – to threaten election integrity, harm public safety, and prey on the most vulnerable members of our society.  
In January, potential primary voters in New Hampshire received a call, purportedly from President Biden, telling them to stay home and “save your vote” by skipping the state’s primary. Steve Holland, “Fake ‘Biden’ robocall tells New Hampshire Democrats to stay home,” Reuters (Jan. 22, 2024), https://www.reuters.com/world/us/fake-biden-robo-call-tells-new-hampshire-voters-stay-home-2024-01-22/. 
  The voice on the call sounded like the President’s, but of course it wasn’t.  Those were voice cloning calls.  The use of generative AI has brought a fresh threat to voter suppression schemes and the campaign season with the heightened believability of fake robocalls. Christine Chung, “The Used Robocalls to Suppress Black Votes.  Now They Have to Register Voters.” The New York Times (Dec. 1, 2022), https://www.nytimes.com/2022/12/01/us/politics/wohl-burkman-voter-suppression-ohio.html; In the Matter of John M. Burkman, Jacob Alexander Wohl, J.M. Burman & Associates, Forfeiture Order, File No. EB-TCD-21-00032652 et al. (2023) (FCC’s assessment of $5,134,000 forfeiture against the perpetrators of 2020 robocall voter suppression scheme for violations of the TCPA). 
  Another example: parents have been scared half to death hearing their child’s voice on the other end of the line, saying they’ve been kidnapped, or need money to get out of trouble. Faith Karimi, “‘Mom, these bad men have me’: She believes scammers cloned her daughter’s voice in a fake kidnapping,” CNN (Apr. 29, 2023), https://www.cnn.com/2023/04/29/us/ai-scam-calls-kidnapping-cec/index.html; Pranshu Verma, “They thought loved ones were calling for help.  It was an AI scam.” The Washington Post (Mar. 5, 2023), https://www.washingtonpost.com/technology/2023/03/05/ai-voice-scam/. 
  In actuality, their children are safe and unaware of the chaos.  Finally, fraudsters continue to target the elderly through “grandparent” scams, but voice cloning adds a frightening new dimension to these attacks on those who may be least able to detect them. Federal Communications Commission (Feb. 1, 2024), ‘Grandparent’ Scams Get More Sophisticated, https://www.fcc.gov/grandparent-scams-get-more-sophisticated. 
  These are real attacks causing harm and confusion, and given their effectiveness, we can assuredly expect more to come.  
The whole of government is rightly focused on whether and how to regulate AI.  President Biden issued a landmark executive order aimed at seizing the promise and managing the risks of AI.  Senator Schumer has convened nine bipartisan AI insight forums, bringing together leading minds to advise Congress broadly on AI’s impact from the workforce to national security.  Legislators have introduced bills addressing the use of AI in dozens of contexts, including election integrity: in fact, a bipartisan group of senators have introduced legislation that would ban the use of AI to generate content falsely depicting federal candidates to influence an election, and comprehensive robocall legislation announced just days ago would double the statutory penalty for calls using AI to impersonate an individual or entity with the intent to defraud.   
This work is critical.  In the interim, agencies that have existing authority to regulate AI should work hard to address harmful uses of the technology.  And they are.  The Federal Trade Commission just finished accepting submissions for its Voice Cloning Challenge – an effort to encourage products, policies, and procedures aimed at protecting consumers from AI-enabled voice cloning harms.  The Federal Election Commission has sought comment on prohibiting the deceptive use of AI in campaign advertisements.
I am proud that the FCC is also stepping in to play its own unique role.  We said it in our November Notice of Inquiry, and today’s Declaratory Ruling makes it clear: the use of voice cloning in telephone calls and texts falls within the FCC’s statutory authority under the TCPA.  The Act prohibits calls using “artificial or prerecorded voice[s]” without consent.  What is voice cloning, if not the use of an artificial voice?  By issuing this item, we’re responding to 26 bipartisan state attorneys general, who last month emphasized to the FCC that “any type of AI technology that generates a human voice should be considered an ‘artificial voice’ for purposes of the TCPA.” Reply Comments of 26 State Attorneys General, CG Docket No. 23-362 at 1 (Jan. 16, 2024).
  
The FCC’s collaborative work with state and local law enforcement is key to investigating and stopping robocall scams.  We must do all we can to aid these partners, so where we can give state attorney general partners the certainty of authority to go after these threats, we absolutely should.  I know we will work hand in glove to enforce the law against bad actors using voice cloning calls to cause harm – and as we’ve seen, the harm can range from financial fraud to threats on the sanctity of our democratic process.  Too much is at stake to stand still.
I want to thank the Chairwoman for her leadership in our robocall efforts and her prompt action on today’s Declaratory Ruling.  My thanks also to the Commission staff who worked on this item.  It has my full support.   


2
STATEMENT OF
COMMISSIONER ANNA M. GOMEZ
Re:	Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, Declaratory Ruling (February 2, 2024)
Artificial Intelligence (AI) technologies can bring both new challenges to the security of our communications and opportunities to combat robocalls and robotexts.  Responsible and ethical implementation of AI technologies is crucial to strike a balance, ensuring that the benefits of AI are harnessed to protect consumers from harm rather than amplify the risks they face in an increasingly digital landscape.  That is why it is important that the Federal Communications Commission issues this Declaratory Ruling.  
Today, the Commission confirms that the Telephone Consumer Protection Act’s prohibition on using “artificial or prerecorded voice” to make robocalls and robotexts without prior express consent of the receiving party includes AI technologies that generate human voices, such as voice cloning.  This Declaratory Ruling is particularly important to prevent fraudsters from using AI to prey on consumers.