Ever fancied hearing Brad Pitt speak flawless Mandarin? That’s the kind of trick Synthesia’s hyper-realistic video avatars can pull off.

The London-based unicorn, which uses generative AI to make videos for corporate training and internal comms, has been a big hit among enterprise users — its customer list includes more than 80% of the Fortune 100, and it hit $100m ARR in April this year.

It seems to be a pretty solid business model — and investors certainly like it. But where will the company take its technology next, how safe are its avatars from deepfake misuse and when might it brave an IPO?

CEO Victor Riparbelli isn’t shy about sharing his views on that front, recently declaring the London Stock Exchange a “complete disaster.”

On this week’s episode, host Amy and senior reporter Kai Nicol-Schwarz put Synthesia’s avatars to the test, dig into its move into agentic AI and ask what, if anything, could go wrong for the AI unicorn.

Sign up to our deeptech and AI newsletter here: https://sifted.eu/members/login?newsletter=deeptech

Take our listener survey here: https://form.typeform.com/to/WbVxsSv7 (T&Cs apply)