ElevenLabs, a startup that uses artificial intelligence software to replicate voices in more than two dozen languages, has raised a new round of financing that values the company – created two years ago – at more than $1 billion.
The company behind deepfake audio, or the famous Deepfake Audio, said it recently raised another $80 million in financing led by venture fund Andreessen Horowitz, with the participation of Sequoia Capital, Smash Capital and SV Angel. The company has raised US$101 million to date. CEO Mati Staniszewski said the latest funding gives his startup a valuation of $1.1 billion.
So far in February, US authorities have received alerts about reports of fake audios that have reached the political scene impersonating both Republican and Democratic leaders.
Fake audio business
Technology investors have allocated funds to startups developing artificial intelligence tools that generate compelling audio and video, taking into account commercial opportunities in advertising and media. But that alarmed researchers concerned about the proliferation of deceptive deepfakes used in scams, pornography and politics, particularly as nearly half of the world’s population will have the option to vote in a national election in 2024. Early last year, ElevenLabs said that some used their technology “for malicious purposes beyond our purpose.”
Staniszewski commented that audio clips that impersonate people without their consent are prohibited and will be removed. The 40-person startup has five people dedicated to moderation and policy work and plans to hire more, he said. “About 99% of the use cases we see are in positive contexts,” she said.
With the new funding, the company plans to expand beyond initial uses in audiobooks and video games into media production, such as dubbing movies or creating fully AI-powered actors, Staniszewski said.
However, this is a thorny ambition. The widespread use of AI in entertainment was a major sticking point in the recent Hollywood strikes. The rise of computer-generated audio, such as ElevenLabs’ tools, has raised concerns that the voices of actors or celebrities could be used without permission or payment. Staniszewski stressed that he wanted to be associated with the entertainment sector.
In addition to the funding announcement, ElevenLabs is introducing a new service that allows certain people to create an AI version of their voice, upload it to a library, and then earn money when it’s used. This year, Staniszewski said, the key question is: “How do we implement this technology safely and together with the industry?”
The startup has spoken to “most” of the big talent agencies and entertainment unions, he added, although he did not elaborate on the conversations. The startup is also working on features to automatically adjust the emotions and intonations of its AI audio. “The technology doesn’t exist yet,” Staniszewski said.
