Commenters Urge FCC To Table Rules On AI

The FCC has proposed rules on “robocalls and robotexts” generated by artificial intelligence, which commenters say the commission should narrow down or delay.

The Federal Communications Commission received pushback in additional comments on its proposed rule on artificial intelligence (AI) communications.

The FCC recently closed public comments and reply comments on its multi-pronged proposed rule (PDF) on AI communications, ACA International previously reported.

The proposals only apply to outgoing calls using AI. There is no proposed regulation for using AI to process or respond to incoming calls. One of the FCC’s primary concerns is the use of AI voice cloning technology to fool consumers.

“The [c]ommission … proposes and seeks comment on measures designed to ensure that its rules keep pace with the fast-developing changes in AI technologies,” according to the Federal Register notice on the proposed rule. “The [c]ommission also seeks to ensure that the [c]ommission’s rules do not hinder the potential benefits that AI technologies can offer, including making telecommunications more readily accessible to individuals with disabilities.”

In initial comments to the FCC (PDF), ACA and other industry trade associations advocated for enforcement actions directed toward bad actors using AI, rather than broad rules that target the consumer-friendly use of the technology.

ACA, America’s Credit Unions, American Financial Services Association, Mortgage Bankers Association and Online Lenders Alliance responded with a call for the commission to “regulate AI technology cautiously and incrementally,” according to the comments.

Stakeholders had the opportunity to submit reply comments to the FCC, in which ACA, America’s Credit Unions, the Mortgage Bankers Association and the Online Lenders Alliance reiterated those concerns and urged the FCC to delay its regulation of AI communications.

“A number of comments point out that regulation of AI is premature given the nascent state of its development and that regulation now is not only unnecessary, but likely will cause confusion and hinder the deployment of AI, all while doing little, if anything, to curb bad actors who will not comply with any consent requirements the commission may adopt,” ACA and the associations state in reply comments (PDF).

The FCC’s notice proposes to address potential harm from AI-generated calls by requiring callers to:

1. Obtain prior consent from consumers to receive an AI generated call, and/or
2. Disclose the use of AI in all instances at the beginning of a voice call, with an exception for use by individuals with disabilities.

However, the “disclosure requirements will not address fraud unless the fraudster complies with them, which will not happen because it will undermine the fraudulent scheme,” according to the comments.

The commenters argue that additional AI consent disclosures are unnecessary, as the Telephone Consumer Protection Act already requires prior express consent for artificial or prerecorded voices. The FCC has confirmed that AI-generated voices fall under this TCPA definition.

Overall, the associations’ comments stress that legitimate callers are already providing required disclosures for calls—such as for debt collection—and provide sufficient notice to consumers.

Callers who cannot obtain a consumer’s specific consent for AI-generated calls as required under the proposed rule will not be able to benefit from the productivity of AI while communicating vital information, such as fraud alerts, school closings or medical appointments.

Additional points from the reply comments include:

• If the commission moves forward with a per-call disclosure requirement, it should be limited to calls where the consumer conversationally interacts with AI and the disclosure should be simple and straightforward.
• Any new AI-related disclosure rules should apply prospectively only. Companies should not be required to obtain additional consent from those that have already consented to receive autodialed or prerecorded or artificial voice calls. If the commission adopts the requirements for the proposed disclosure when making an AI-generated call, it should confirm that existing exemptions from consent apply.
• As it continues to assess ways to combat illegal calls, the commission should urge industry to move quickly in adopting standardized technology and methods to verify the identity of the caller.