Former Plaud executive Peter Mo has secured $26 million in funding for Ambi, an AI application designed to deliver real-time intelligence by monitoring users continuously throughout the day, according to a company announcement Monday. The funding came across multiple rounds led by Luminous Ventures, which has previously backed Meituan and Unitree Robotics, alongside original design manufacturer Huaqin. Ambi is launching applications for iPhones, Macs, and Apple Watch that function like meeting notetakers but allow users to enable all-day listening to capture context continuously.

The startup has assembled a workforce exceeding 50 employees spread across locations in San Francisco, Beijing, Shanghai, and Shenzhen, and is currently recruiting for go-to-market positions and AI engineers focused on speech-recognition models. Ambi's AI agent aims to proactively offer suggestions or surface relevant knowledge during meetings by drawing on users' calendars, meetings, knowledge bases, and web data. The company's forthcoming desktop application will monitor the work users perform and retain important segments in its memory to deliver more personalized recommendations. A planned hardware device will incorporate a screen, speaker, camera, and microphones to capture context and display information.

"The way we interact with intelligence is limited," Mo, the company's co-founder and CEO, told TechCrunch. He explained that intelligence often remains confined to screens, and hardware like Plaud requires recordings to transfer to phones, upload to the cloud, and undergo processing before delivering intelligence. The company wants to differentiate itself by enabling its AI agent to begin work during meetings themselves, going beyond what meeting notetaking companies currently offer by automating tasks after calls conclude. Ambi allows users to start tasks on one device and continue them on another device.

Mo acknowledged that ambient data capture is turned off by default and users can disable it at any time, though the startup didn't specify what encryption method it employs despite stating it encrypts data and doesn't store recordings. He classified disclosure as an industry-wide challenge, noting that the company has attempted to place disclosure information everywhere possible, but can't prevent users from concealing capture if they choose to do so. The startup's launch arrives amid mounting concerns about AI devices lacking sufficient indicators to show they're capturing data on people surrounding the user, with some meeting notetakers facing lawsuits because their technology recorded conversations and data without consent. Apps like Littlebird or Minimi already perform functions similar to what Ambi's desktop application plans to offer, while Apple's recently released Audio Intelligence feature, which transcribes speech continuously, has sparked privacy concerns.

The company represents a wave of startups building continuously recording apps despite obvious privacy and data security concerns, armed with the conviction that the market for personalized assistance is large enough to overcome such worries. These companies operate on the core promise that AI improves its ability to assist with tasks and make appropriate suggestions the more it knows about a person, which tech companies say requires apps and devices to listen, watch, or record users' activity constantly and gather all available context. Mo said users must inform others about ambient capture, acknowledging the responsibility for declaring ambient capture still falls on users. The challenge for companies pursuing ambient intelligence will be whether consumers trust them enough to grant constant access to their conversations and activities, particularly when third-party consent remains largely unenforceable.