The top talent behind yet another promising AI startup has been gobbled up by an incumbent. As part of a new licensing agreement, Google DeepMind is bringing on the CEO and several of the top engineers at voice AI startup Hume AI.
Wired was the first to report the news.
CEO Alan Cowen and about seven other engineers will work with DeepMind to improve Gemini’s voice features, according to Wired.
What’s left of Hume AI will continue to supply its technology to other AI firms. No financial details of the deal were shared, but Andrew Ettinger, an investor and tech executive who joined the Hume AI team last week and has taken over as CEO, told TechCrunch that Google has a “non-exclusive right to certain technologies, and we’ll be infusing that into their processes.”
He added that Hume AI would be releasing new models in the coming months, and is on track to bring in $100 million in revenue this year.
The Hume acqui-hire is the latest example of a leading AI firm scooping top talent off the market and skirting regulatory scrutiny by acquiring a startup’s team rather than the company outright. Last year, Google acquired viral AI coding startup Windsurf’s CEO and other top researchers, and OpenAI has acquired several startup teams in recent months, including Convogo and Roi. The Federal Trade Commission recently said that it would take a closer look at such deals.
The deal also shows that voice is becoming the next frontier in AI.
Techcrunch event
San Francisco
|
October 13-15, 2026
Hume AI’s secret sauce is its model’s ability to understand a user’s emotions and mood based on their voice. In 2024, the startup launched its Empathetic Voice Interface, a conversational AI with emotional intelligence. Hume AI has raised close to $80 million to date according to PitchBook.
But Hume AI isn’t the only company working on voice-focused models. Google has been steadily improving its Gemini Live feature, which allows a user to have conversations with the chatbot. Last month, Google released a new native audio model for the Live API that improved the model’s ability to “handle complex workflows,” per the Gemini API release notes.
Others in the industry are investing big into voice capabilities as well. OpenAI is reportedly preparing to overhaul its audio models in preparation for its audio-first personal device, created with Jonny Ive’s io, to launch this year. Recent leaks suggest the device could be a form of earbuds.
Last year, Meta also accelerated its AI audio push by acquiring startup Play AI. The Facebook-maker’s Ray-Ban smart glasses are increasingly relying on voice and audio capabilities for tasks like helping you hear conversations in noisy rooms and enabling hands-free control for calls, texts, music, and photos.
“Voice is the only acceptable input mode for wearables,” investor Vanessa Larco told TechCrunch. “This acquisition will only accelerate the need for voice apps.”
Demand continues to increase for voice capabilities. Earlier this month, ElevenLabs, the AI voice generation startup, said it crossed $330 million in annual recurring revenue.
This article has been updated with comment from Hume AI and to clarify details of Google’s licensing deal with the startup.
