Artificial Intelligence (AI)-generated robocalls may trick some consumers into thinking they are being called by a human being, but the Federal Communications Commission clarified in a recent AI Declaratory Ruling that it will not be fooled. Moving forward, all AI-generated robocalls will be treated as artificial or prerecorded voice calls for purposes of the Telephone Consumer Protection Act (TCPA) and will require a called party’s prior express consent. The AI Declaratory Ruling reflects a first step by the FCC in crafting a new record on AI’s implications for consumers’ rights under the TCPA. That record began last November when the FCC released an AI-focused Notice of Inquiry (NOI), which sought industry and stakeholder comments on the potential benefits and risks of AI for consumers. By confirming that AI technologies used in robocalls are artificial or prerecorded voice calls under the TCPA, the FCC hopes to stem the rapid proliferation of AI-generated robocall scams, including popular “deep-fake” or voice cloning scams that solicit money by mimicking the voices of popular celebrities or even family members.
 
Admittedly, the FCC’s conclusion that AI-generated voice calls are “artificial” seems, at face value, axiomatic. But prior to the AI Declaratory Ruling, there was at least a colorable argument that AI-generated human-sounding voices capable of engaging in a live, interactive conversation was sufficient to qualify for the TCPA’s longstanding prior consent exception for live calls from human beings. That theory has been shut down, at least under current AI technologies. 
 
Defining AI Technologies

In its NOI, the FCC questioned “whether it is necessary or even possible to determine at this point whether future types of current AI technologies fall within the TCPA’s existing prohibitions on artificial or prerecorded voice messages.” In the AI Declaratory Ruling, the FCC seemingly hedged its bet that future advances in AI could cause the FCC to change its position. It declared that the TCPA prohibits “current AI technologies that resemble human voices and/or generate call content using a prerecorded voice,” which includes voice cloning and other contemporary AI-generated messages. 

In limiting the AI Declaratory Ruling to current AI technologies, the FCC sought to avoid an overly broad determination that might discourage advances in AI voice-related technologies in ways that benefit consumers. FCC Chair Jessica Rosenworcel also noted that this agency determination may facilitate collaboration between state attorneys general and the FCC in identifying and responding to AI-related robocall violations. 
   
No “Functional Equivalent” for Live Agents

While the FCC left the door open a crack by tying its determination to current AI technologies, it kept the door from swinging open widely by declaring there is no exception from the prior consent requirement for robocalls that provide the “functional equivalent” of live, interactive human voice calls. Drawing on a 2020 declaratory ruling that considered callers’ use of “soundboard technology” to send messages to consumers (i.e., when an agent plays multiple short, prerecorded messages in response to consumer cues), the AI Declaratory Ruling clarifies that “the TCPA does not allow for any carve out of technologies that purport to provide the equivalent of a live agent.”  
 
Future Actions

The AI Declaratory Ruling does not address all of the open issues raised in the NOI. For example, the FCC’s robocall rules currently require that callers using an artificial or prerecorded voice provide clear information identifying the entity responsible for initiating the call at the beginning of the call. In the NOI, the FCC sought comments on amending this rule to require a type of “digital watermark” indicating whether a voice is AI-generated or otherwise requiring callers to disclose on the call that they that are using AI technologies to emulate a real human being. It is possible that, in light of the recent high-profile case of robocalling in the New Hampshire primary that relied on AI-generated technology to spoof the voice of President Biden, the FCC wanted to issue this clarification now, in advance of the run-up to the general election in November, while it continues to evaluate next steps on the issues raised in the NOI. 

© 2024 Perkins Coie LLP