Jeff Geerling Accuses Elecrow of Using an AI Voice Clone in Product Videos

# Jeff Geerling Accuses Elecrow of Using an AI Voice Clone in Product Videos

*Event date: 2024-09-22*

By IO Digest Desk

A blog post published around September 22, 2024 put a personal and commercial face on one of AI’s most unsettled questions: what happens when a recognizable voice can be copied cheaply enough to appear in marketing material without a clear agreement from the speaker. In the post, creator Jeff Geerling says electronics company Elecrow published tutorial videos narrated in a voice that sounded so much like his that viewers could reasonably assume he had participated. He says he never recorded the lines used in those videos.

Geerling’s account is careful in one respect and forceful in another. He says he cannot definitively prove how the audio was produced, but he is “pretty sure” Elecrow fed his YouTube material into an AI voice-cloning system and then used the result to narrate multiple promotional series, including videos about ESP32 and RP2040 hardware. He frames the issue not only as imitation but as commercial appropriation: a company using a familiar creator voice to promote products without consent.

The accusation lands differently because Geerling says he did not have a hostile prior relationship with the company. He notes that he had reviewed an Elecrow product in the past and that the company makes electronics and Raspberry Pi accessories. More importantly, he says Elecrow knew his work well, pointing to more than 43 emails exchanged with five Elecrow marketing representatives from 2020 through 2024, including a paid-partnership inquiry sent on April 2, 2024. That history does not prove intentional cloning, but it does make mistaken resemblance harder to dismiss as random coincidence.

The post also captures the legal and platform uncertainty that still surrounds synthetic voice use. Geerling says he is not aware of a direct legal precedent specifically about unauthorized AI voice cloning, though he notes there is precedent against using someone’s voice in commercial works without permission and references Midler v. Ford. He is similarly uncertain about whether non-consensual voice cloning clearly violates YouTube’s terms. Those remarks matter because they show how much of the present conflict is happening in a gray zone between obvious ethical discomfort and still-forming enforcement mechanisms.

Rather than start with legal action or platform takedowns, Geerling says he emailed Elecrow requesting that at least two tutorial series be removed and asking whether the company intentionally trained a voice on his content. He presents that as a first step motivated partly by caution: he says he is not 100 percent certain the imitation was intentional, even if he considers the similarity overwhelming. The post was later updated to note that Elecrow’s CEO had responded, though the evidence packet for this assignment does not include the substance of that reply.

What the article clearly does establish is the reputational risk synthetic media now creates for brands and creators alike. If audiences believe a familiar person is endorsing a tutorial, product, or campaign when that person never agreed to participate, voice cloning stops being a novelty and becomes a trust attack.

Geerling’s broader argument is blunt and easy to understand. Companies that want a creator’s likeness or vocal identity should hire the creator or a voice actor instead of simulating the relationship after the fact. As generative tools continue to lower the cost of plausible imitation, that norm may prove as important as whatever legal rule eventually follows.