Meta’s latest acquisition of WaveForms AI may appear to be just another addition to its growing list of AI-related deals, but a closer look reveals a deliberate move to strengthen its position in expressive and emotionally aware voice technology. Unlike previous acquisitions focused on mainly technical improvements or scale, this acquisition seems to target a specific gap in the AI voice space that is the ability to convey emotions naturally.
WaveForms, founded in late 2024, quickly became a name to watch in the AI audio space. Within eight months of its launch, the startup secured $40 million in funding from Andreessen Horowitz, reaching a pre-money valuation of around $160 million. Its mission is ambitious: to build what it calls “Emotional General Intelligence” in audio AI, aiming for a “Speech Turing Test” where a listener cannot distinguish between human and machine.
The company’s founders, Alexis Conneau and Coralie Lemaitre, bring considerable expertise to the table. Conneau, a former Meta and OpenAI researcher, contributed to the development of GPT-4o’s voice capabilities, including systems designed to shift tone in real time. Lemaitre, with experience at Google, has a background in advanced speech synthesis and prosody modeling. Both have now joined Meta’s Superintelligence Labs, making this acquisition as much about securing talent as it is about acquiring technology.
