Cisco says it will acquire BabbleLabs, which uses AI to remove unwanted noise and detect human speech, to integrate into Webex and other collaboration offerings
Context & Ripple Effects
This is the third voice-AI acquisition Cisco has bolted onto its collaboration suite: it bought conversational-AI firm MindMeld in 2017 and shipped the Spark Assistant on top of it, then picked up Voicea and its Eva meeting assistant a year ago. BabbleLabs sits one layer lower in the stack — not meeting semantics but raw audio quality — and only recently emerged from stealth with a $14M Series A aimed at improving speech quality and accuracy across third-party voice apps.
The strategic logic differs from those earlier deals: instead of adding another assistant layer alongside Voicea's Eva, Cisco is buying signal hygiene itself — noise removal and human-speech detection — which becomes an invisible differentiator baked into every Webex call rather than a feature users opt into.
First-order effects
- Webex users get AI noise suppression and speech detection built into calls, and Cisco gains a quality lever it controls end-to-end rather than depending on device or client defaults.
Second-order effects
- BabbleLabs had been positioning its technology for other voice-app developers; absorption into Webex removes it from the merchant market, pushing rival collaboration platforms to source or build equivalent audio cleanup for their own clients.
Third-order effects
- If Cisco keeps folding acquired speech AI into the suite — MindMeld's assistant, Voicea's transcription, now BabbleLabs' audio layer — standalone voice-AI startups increasingly exit via acquisition into platform bundles, and call quality becomes a baseline expectation competitors must meet by default.
The trend: Collaboration platforms are consolidating speech-AI startups feature by feature, turning audio quality and voice intelligence from purchasable add-ons into embedded suite defaults.