On November 16, 2023, the Federal Communications Commission (FCC) released a Notice of Inquiry (NOI) to seek information on the implications of emerging artificial intelligence (AI) technologies on robocalling, and the Federal Trade Commission (FTC) released a voice cloning initiative aimed at developing solutions to protect consumers from AI-enabled voice cloning harms. The items further demonstrate federal regulators’ increased focus on AI technologies, and their potential benefits and risks in the context of robocalls and robotexts.

FCC Notice of Inquiry

The NOI seeks information on the implications of emerging AI technologies as part of the FCC’s ongoing efforts to protect consumers from unwanted and illegal telephone calls and text messages under the Telephone Consumer Protection Act (TCPA). Comments are due on December 18, 2023, and reply comments are due January 16, 2024.

The NOI seeks comment on how to define AI in the TCPA context.

The NOI seeks comment on how to define AI for purposes of fulfilling the FCC’s statutory responsibilities under the TCPA. The FCC asks about various definitions of AI and references how the federal government has started to address emerging AI technologies, including through a recent Executive Order. The NOI also explains that the TCPA’s existing prohibition on “artificial” voice messages encompasses current AI technologies that generate human voices, and asks whether to define AI technologies in a more specific way to consider how AI can function as the equivalent of a live agent when interacting with consumers. The FCC further questions whether AI should be defined differently from other existing technologies, such as Interactive Voice Response (IVR) systems and voice cloning.

The NOI focuses almost exclusively on considerations under the TCPA.

Beyond the focus on defining AI in the TCPA context, the NOI heavily focuses on the TCPA and asks whether the FCC’s legal authority under the TCPA is sufficient to enact rules to ensure that AI technologies do not erode consumer protections under the TCPA or other consumer protections statutes. The NOI mentions other potentially relevant legal frameworks, like CAN-SPAM, but does not reference the TRACED Act.

The FCC seeks comment on whether existing legal authorities under the TCPA or other consumer protections statutes provide it with sufficient authority to enact rules to ensure that AI technologies do not erode consumer protections. The FCC also asks about its authority under the TCPA to make “technical and procedural standards for systems that are used to transmit any artificial or prerecorded voice message via telephone.”

The FCC requests information on regulating AI under the TCPA in a number of ways, including for both prerecorded and artificial voice calls and autodialer calls. Importantly, the FCC seeks comment on whether the use of AI could qualify as an “autodialer” under the TCPA’s current definition and whether this should further impact how the FCC defines AI in the robocall context.

The NOI seeks comment on the benefits of AI in the robocalling and robotexting context.

The FCC also seeks comment on the benefits of AI technologies in the TCPA context. Specifically, the NOI seeks information on how AI can be used to: (1) protect consumers from unwanted robocalls and robotexts; (2) improve the FCC’s ability to enforce the TCPA; and (3) improve accessibility.

The FCC asks whether AI can be used to ensure called parties only receive robocalls and robotexts on subjects they are interested in—including whether AI could improve the ability of networks to detect unwanted or fraudulent traffic and stop unwanted calls before they reach consumers. In instances where providers block calls based on analyzed call content, the FCC seeks comment on the possible privacy concerns and how to address them.

The FCC also requests information on whether voice service providers and third-party analytics companies already use AI to comply with the TCPA. The FCC asks whether AI technology can process do-not-call requests, ensure that only parties who have provided consent are called, keep track of numerical limits for calls to specific telephone numbers, and avoid calls to emergency lines or other wireless telephone numbers prohibited under the TCPA.

To improve accessibility, the FCC asks whether AI technologies can improve the ability of persons with disabilities to communicate with a called party—the FCC questions whether AI can work with telecommunications relay services (TRS) or facilitate pausing calls until a TRS Communications Assistant successfully connects with a TRS user. The FCC seeks comment on how current and future AI systems may interact with accessibility tools that generate or translate speech.

The NOI also seeks comment on the risks of AI in the robocalling and robotexting context.

The NOI acknowledges that AI technologies could also be used to make legal, fraudulent, or otherwise unwanted robocalls and robotexts or avoid detection by analytics systems that block such messages. The FCC asks whether AI may inadvertently block legitimate calls and messages and how the FCC or industry could take steps to prevent such blocking. The NOI also seeks comment on how AI technology could increase costs for industry stakeholders seeking to detect and reduce high volumes of robocalls and robotexts in their networks.

In the context of impersonation calls, the NOI asks whether the FCC should consider ways to verify the authenticity of legitimately generated AI voice or text content through watermarks, certificates, labels, signatures, or other methods. The FCC seeks input on whether it could use its authority under section 227(d)(3), or any other authority, to require the use of such “digital watermarks” to indicate whether a voice on a robocall is generated by AI. To further identify AI-generated or AI-assisted robocalls, the FCC asks whether a disclosure—similar to identification requirements under the TCPA—should be implemented to further disclose the use of AI for robocalls. The FCC also seeks comment on a variety of miscellaneous issues, such as whether AI could introduce harmful bias, disrupt elections, or make false emergency robocalls, and whether AI-assisted robocalls and robotexts should be treated differently from traditional robocalls and robotexts.

To address potential areas of concern for AI technology, the NOI asks how to identify the root causes of AI-driven robocall or robotexts scams—including whether to solicit information from industry regarding the type of AI technologies used in scams, either on a regular basis or in connection with investigations. The NOI considers whether to inquire about the purpose for which AI technology was developed (i.e., for legal uses versus unlawful uses) and how to share the information about fraudulent uses of AI with other government agencies tasked with addressing malicious uses of AI in other contexts.

The FCC asks how to handle liability for AI developers.

Lastly, the FCC asks whether developers of AI—and not just the robocallers using AI technologies—should be held accountable if their systems are designed to operate in a way that does not comply with the TCPA or other related consumer protection statutes. The FCC seeks comment on what else it should consider to ensure that the TCPA’s protections afforded to emergency lines (i.e., 911 lines, emergency lines of a hospital, poison control center, or law enforcement) are not undermined by AI technologies.

FTC Voice Cloning Challenge

Separately, the FTC announced the launch of its Voice Cloning Challenge (Challenge), to encourage the development of different approaches—ranging from products to policies and procedures—to protect consumers from AI-enabled voice cloning (the creation of an artificial simulation of a person’s voice). Like the FCC NOI, the FTC cites the significant risk for voice cloning technology to perpetuate fraud. The FTC notes that the Challenge is part of a broader strategy to tackle the risks posed by voice cloning and other AI technology, and that the agency “will be using all of [its] tools” to address the issue, including enforcement and rulemaking.

The FTC issued detailed rules for the Challenge, which permits individuals, for-profit companies, non-profit organizations, and others to submit ideas aimed at protecting consumers from AI-enabled voice cloning harms such as fraud and the broader misuse of biometric data and creative content. Submissions to the Challenge must—at a minimum—address one of the following voice cloning intervention points: (1) methods to limit the use and application of voice cloning software by unauthorized uses; (2) methods to detect cloned voices or the use of voice cloning technology; or (3) methods to check after the fact if audio clips contain cloned voices.

The FTC rules for the Challenge explain that submissions should be anonymous and should not contain information revealing the participant’s identity. Submissions to the challenge can be submitted online from January 2, 2024 to January 12, 2024, and the FTC expects to announce the results in early 2024.

[View source.]