FCC Proposes Massive Fine for Illegal AI Robocalls

Table of Contents

Questionable Motives

Far Reaching Results

Two days before January’s 2024 New Hampshire primary election, thousands of residents of the Granite State received illegally spoofed AI robocalls. While this seems to occur with depressing regularity in every state these days, these calls were decidedly different.

These calls were allegedly placed by a political operative named Steve Kramer, who used generative artificial intelligence (AI) to mimic the voice of President Biden to convince voters to refrain from voting in the primary election. To ensure the calls were answered, he also spoofed the number of a prominent New Hampshire Democratic Party member, thus reinforcing the credibility of the calls.

The calls were placed using the services of Lingo Telecom, which is accused of disregarding STIR/SHAKEN standards to further strengthen the legitimacy of the calls. For those unfamiliar with them, Secure Telephone Identity Revisited (STIR) and Signature-based Handling of Asserted Information Using toKENs (SHAKEN) establish a framework of interconnected standards that allow for the authentication and verification of caller ID information for calls carried over VoIP networks.

When a call travels via VoIP networks, STIR/SHAKEN enables the originating carrier to attest to the authenticity of the caller ID information accompanying the call. Each downstream carrier then passes on the attestation level provided by the originating carrier before it reaches its final destination. Lingo Telecom signed the calls with an “A” attestation- the highest level of trust available- without undertaking any effort to verify that the caller ID information was accurate.

This made it far more likely that the calls would be answered by their intended recipients.

Questionable Motives for AI Calls

Sometime after Kramer was identified as the mastermind behind the January AI calls, he gave an interview with CBS news in which he dubiously asserted that the sole motive behind the entire scheme was to demonstrate the need to regulate AI to the FCC. He went on to claim that the $500 it cost him to place the calls was “money well spent,” as the stunt generated $5 million in free publicity.

In explaining why it signed the calls with an A attestation, Lingo claims that it relied solely upon the contractual statements provided by the agencies Kramer used to place the calls with Lingo, which consisted of a single page form with little to no substantive information.

Far Reaching Results

Although Kramer’s shaky explanation for initiating the scheme strains credulity, few can deny it did indeed get the attention of the FCC, the New Hampshire Attorney General, and just about everyone else, and on February 8, 2024, the FCC unanimously adopted a Declaratory Ruling officially decreeing an AI-generated voice to be “artificial” as that term is used in the Telephone Consumer Protection Act (TCPA).

Unfortunately for Mr. Kramer, the AI regulation he claims was the sole reason behind placing the calls was not the only consequence of his scheme. As detailed in a prior article, in March a federal lawsuit was filed in New Hampshire against Kramer and the companies he used to place the calls alleging TCPA and Voting Rights Act violations. The New Hampshire Attorney General also initiated a criminal complaint, against Kramer, charging him with 13 felony counts of voter suppression and 13 misdemeanor counts of impersonating of a candidate. And to add yet another insult to these injuries, on May 24, 2024, the FCC announced that it intended to fine Kramer $6 million for his role in the absurd stunt and proposed a $2 million fine for Lingo Telecom.

While this matter is far from concluded, it is clear that using AI to interfere with an election has generated immense consequences. As new rules continue to be enacted and old ones clarified, companies must take note of these events and be extremely wary when using AI to contact consumers, and telecom providers must take active efforts to comply with STIR/SHAKEN standards.