- ElevenLabs, a startup focused on voice synthesis, raised $80m to reach unicorn status.
- This comes as the company adds new features, like its dubbing studio for content production.
- However, fears surrounding the potential for ElevenLabs to create misinformation are rising.
ElevenLabs, a company that uses AI to generate natural, lifelike voices, has raised a massive $80m in its Series B funding round.
The ElevenLabs platform allows users to create AI voices in numerous languages and accents, complete with a wide range of emotional and intonational variations and nuances.
According to a press release, the company’s technology has been widely adopted, with users generating over a century’s worth of audio.
The company has grown rapidly, with an estimated 41% of Fortune 500 companies currently using its technology.
Piotr Dąbkowski, ElevenLabs’ CTO, described the impact of this new funding as follows: “Our team’s dedication has already made a lasting impact on voice AI. This new funding enables us to take on even bigger challenges and focus on maintaining our competitive edge in research and product development.”
ElevenLabs expands offerings
In addition to its Series B funding announcement, ElevenLabs is introducing several innovative product developments:
- Dubbing Studio: This new platform enables users to dub entire movies, manage transcripts, translations, and timecodes, offering greater control over content production. It complements the existing AI dubbing feature, which provides automated video localization in 29 languages.
- Voice Library Marketplace: A unique marketplace where users can monetize AI versions of their voices. Users can create, verify, and share their professional AI voice replicas. Original creators are compensated when these verified voices are used. The marketplace is already generating income for a select group of users.
- Mobile App Reader (Early-Preview): An app that converts text and URLs into audio, facilitating content access on the move. It will initially be free for three months.
ElevenLabs provokes discussions surrounding deep fakes
As this technology evolves, concerns are mounting about its potential misuse.
Last year, ElevenLabs revealed on X that certain actors had exploited their technology for malicious purposes, though specifics were not disclosed. Such deep fakes, often featuring celebrities or politicians, can be used to spread fake news, defame, or create fake pornography.
In the context of the upcoming 2024 US election, GlobalData analyst Emma Christy highlighted the risks associated with audio deep fakes:
“Audio deepfakes are troubling as it is easier and cheaper to replicate voice without the corresponding video, and they are difficult for even technology to detect. Many people will be unable to discern deepfake audio from reality, with catastrophic implications for countries holding elections this year.”
Christy further notes that the proliferation of open-source AI and the increasing prevalence of convincing audio deepfakes could lead to a scenario where voters might distrust legitimate material, a phenomenon known as the liar’s dividend.
Deep fakes will go on as one of AI’s leading risks, and compliance demands for companies like ElevenLabs will increase. However, it’s difficult to see them slowing down, for now, at least.