The FCC fined Lingo Telecom $1 million for carrying the AI Biden deepfake robocalls that told New Hampshire voters to stay home

Verification status: CHECKING — handed to the checker, not yet graduated. NOT verified, NOT green. This is a telecom-enforcement record about an AI deepfake used for election interference, an honest-negative, not a client testimonial. No green badge is sought; the war-room never sets verified.

The problem

The 2024 election was the first in which a cloned voice was weaponised at scale over the phone network. Two days before the New Hampshire 2024 presidential primary, “spoofed robocalls carried a deepfake audio recording of President Biden’s cloned voice telling prospective voters not to vote in the upcoming primary” (source). The calls were “directed by a political consultant named Steve Kramer in an attempt to interfere in the 2024 New Hampshire primary election” (source). They reached voters not because the AI was sophisticated, but because the phone network trusted them: the calls were transmitted through Lingo Telecom, a “voice service provider that transmitted spoofed robocalls that used generative AI voice cloning technology to spread disinformation in connection with a presidential primary election in New Hampshire” (source).

What was built

The technical failure was one of attestation, not audio. Under the FCC’s STIR/SHAKEN caller-ID rules a carrier vouches for who is really placing a call, and Lingo vouched at the highest level for calls whose number was spoofed. In the settlement Lingo agreed to apply “an A-level attestation, which is the highest level of trust attributed to a phone number, only to a call where Lingo Telecom itself has provided the caller ID number to the party making the call” (source) — a term that only makes sense as a remedy because it had not done so here. CyberScoop, reporting on the consent decree, confirmed independently that “all those calls were signed with ‘A-level attestations’ — a designation that indicates the signing provider has a direct authenticated relationship with the customer, and the customer has the right to use the telephone number in the caller ID field” (source), while the number displayed had in fact been spoofed to that of Kathy Sullivan, a former state Democratic official running a write-in campaign for Biden (source). Sources disagree on the volume: per CyberScoop, “Lingo told the FCC in May in response to a subpoena that they transmitted 3,978 calls to New Hampshire voters on Jan. 21, 2024. In a separate section, authorities said approximately 9,581 calls were sent to New Hampshire voters in the incident” (source), and that conflict is left unresolved here rather than averaged into a single figure.

The outcome

The FCC announced that “the company will pay a $1 million civil penalty and implement a historic compliance plan — the first of its kind secured by the FCC — that will require strict adherence to the FCC’s STIR/SHAKEN caller ID authentication rules, including requirements that the company abide by Know Your Customer (KYC) and Know Your Upstream Provider (KYUP) principles” (source). That figure was a reduction: the settlement “resolves the proposed fine, formally called a Notice of Apparent Liability for Forfeiture, that the Commission approved in May” (source), and CyberScoop reported that “the fine represents half the amount the FCC was originally seeking,” namely when “the agency proposed a $2 million enforcement action” (source). FCC Chairwoman Jessica Rosenworcel framed the action around trust in the network: “Every one of us deserves to know that the voice on the line is exactly who they claim to be. If AI is being used, that should be made clear to any consumer, citizen, and voter who encounters it” (source). Enforcement Bureau Chief Loyaan A. Egal said the settlement “sends a strong message that communications service providers are the first line of defense against these threats and will be held accountable to ensure they do their part to protect the American public” (source).

The action against the carrier was separate from the action against the man who ordered the calls. The FCC noted that it “also issued a separate enforcement action in May against Mr. Kramer,” that “the enforcement action against Mr. Kramer for the above-referenced election interference robocalls remains pending,” and that “he faces a proposed $6 million fine” (source); CyberScoop reported that “New Hampshire authorities charged him with 13 felony and 13 misdemeanor counts of impersonating a political candidate” (source).

Weakest load-bearing source. The load-bearing evidence is the Tier-1 primary: the FCC’s own press release announcing the settlement, from which every dollar figure and every FCC quotation above is taken verbatim. Its honest limit is that it is the agency’s summary of the action, not the signed Consent Decree itself, and it records a settlement the carrier consented to, not a court’s finding of liability. The Tier-2 corroboration is CyberScoop, an independent security and policy newsroom rather than a primary docket; it independently confirms the $1 million penalty, the $2 million proposed figure, the A-level attestations, the STIR/SHAKEN failures and the Kramer charges, but the exact call count rests on CyberScoop alone and is internally contradictory (3,978 vs ~9,581), so it is presented as an open conflict, not a verified number. The one figure that would settle it — the operative call count in the signed Consent Decree (FCC EB, DA-24-790) — has not been quoted from the primary document and is recorded as the path to green, not asserted here.

How this was verified. Method: every figure is quoted verbatim from the U.S. Federal Communications Commission’s press release, “FCC Settles Case Against Provider That Transmitted Spoofed AI-Generated Robocalls For Election Interference In New Hampshire” (21 August 2024 — Tier 1, primary; the FCC is the enforcing agency). It was retrieved this session from the archived press-release PDF (docs.fcc.gov DOC-404951A1; Wayback 20260821131613) because ftc.gov and fcc.gov refuse automated fetches, and the cleaned text is saved to sources/fcc-press-release-DOC-404951A1-2024-08-21.txt. Independent corroboration is CyberScoop, “Telecom behind AI-powered Biden robocall agrees to $1 million FCC fine” (Tier 2; Wayback 20260521065751; saved to sources/cyberscoop-2024-08-21.txt), which reproduces the penalty, the proposed figure, the attestations, the STIR/SHAKEN failures and the Kramer charges firsthand. The exact number of calls is left as a shown conflict, not merged. No confirmation was sought from Lingo Telecom or Kramer; the badge never depends on a subject confirming its own conduct. Verified 2026-08-26 (checking round).

Sources

1. U.S. Federal Communications Commission · FCC Settles Case Against Provider That Transmitted Spoofed AI-Generated Robocalls For Election Interference In New Hampshire · 21 August 2024 · https://docs.fcc.gov/public/attachments/DOC-404951A1.pdf — Tier 1 (primary; the enforcing agency’s own release announcing the $1 million settlement and compliance plan; landing page https://www.fcc.gov/document/fcc-settles-spoofed-ai-generated-robocalls-case; archived Wayback 20260821131613, cleaned text saved to sources/fcc-press-release-DOC-404951A1-2024-08-21.txt).
2. CyberScoop · Telecom behind AI-powered Biden robocall agrees to $1 million FCC fine · 21 August 2024 · https://cyberscoop.com/telecom-behind-ai-powered-biden-robocall-agrees-to-1-million-fcc-fine/ — Tier 2 (independent security/policy newsroom; corroborates the $1 million penalty, the $2 million proposed figure, the A-level attestations, the spoofed Kathy Sullivan number, the STIR/SHAKEN failures and the Kramer charges; the call-count figures 3,978 and ~9,581 rest on this source; archived Wayback 20260521065751, saved to sources/cyberscoop-2024-08-21.txt).

Related case files

• The FTC’s $930,000 settlement with Cox Media Group over an “Active Listening” AI ad service that never listened — a US federal regulator pricing the gap between an AI marketing claim and reality, the same enforcement genre in the same window.
• The Italian Garante’s warning over an AI deepfake of anchor Enrico Mentana — a regulator acting on a generative-AI clone of a named public figure, the same deepfake-harm axis from a data-protection angle.
• The FTC’s action banning Air AI from marketing business opportunities over deceptive AI claims — the same posture of a federal agency holding a company accountable for an AI capability that was oversold rather than delivered.