Former Plaud executive secures $26,000,000 for AI application that listens to you in the background

Meta description: Ambi raised $26 million to launch always-on AI apps for iPhone, Mac and Apple Watch, aiming to capture context while raising privacy questions.

Tags: AI assistants, ambient computing, privacy, Ambi, workplace productivity

Featured image: Abstract technology image showing connected data streams and privacy shields, no real people, no company logos, no text.

Ambi has raised $26 million across multiple funding rounds led by Luminous Ventures and Huaqin to build always-on AI assistant apps for iPhone, Mac and Apple Watch, a move that pushes personal AI further into continuous listening while placing privacy disclosure back in focus.

Background: The Promise and Problem of Context-Aware AI

The pitch for personalized AI assistants is simple: the more a system knows about a person, the better it can help with tasks and suggestions. For that to work at scale, tech companies argue, apps and devices may need to listen, watch, or record activity continuously and gather broad context.

That promise comes with a direct trade-off. Systems that collect data all day can capture not only the user, but also bystanders, colleagues, clients, and private conversations. Even so, startups are building around the belief that demand for personalized assistance is large enough to justify the risk.

Former Plaud executive Peter Mo is pursuing that model with Ambi. The company announced on Monday that it has raised $26 million across multiple funding rounds led by Luminous Ventures, which has backed companies such as Meituan and Unitree Robotics, and Original Design Manufacturer Huaqin.

Development Process: Funding, Apps, and Hardware Plans

Ambi is releasing apps for iPhone, Mac, and Apple Watch. They can function like meeting note-takers, but users can choose to keep them listening so the assistant can capture context throughout the day.

The startup is trying to move beyond the usual pattern in meeting note-taking software. Many products automate post-call tasks using summaries and action items collected during conversations, then push those outputs into other systems through many connectors. Ambi wants its agent to begin helping during the meeting itself.

Its stated promise is that the AI agent can proactively give suggestions or surface relevant knowledge while the user is in a meeting, drawing on the user’s calendar, meetings, knowledge base, and web data. Users can also start a task on one device and continue it on another.

Mo’s broader point, as described to TechCrunch, is that intelligence is often trapped on the screen. With hardware such as Plaud, users may need to wait for recordings to transfer to a phone, upload to the cloud, and process before useful insights appear. Ambi’s goal is to make AI available anytime and anywhere.

The company also plans a desktop app that can look at the work users do and store important parts in memory for more personalized suggestions. Apps such as Littlebird and Minimi already use similar approaches. Ambi is additionally planning its own hardware device with a screen, speaker, camera, and microphones to capture context and display information.

Latest Progress: Privacy Questions and Current Stage

Ambi’s launch arrives while scrutiny of AI devices is increasing, especially for tools that may capture data from people nearby without obvious indicators. It remains unclear whether Ambi’s apps or its upcoming device will include enough signals to alert third parties when their conversations are being recorded, particularly if the user does not disclose that Ambi is active.

Some meeting note-taking products have already been sued after recording people’s conversations and data without consent. The recent release of Apple's Voice Intelligence feature, which transcribes speech continuously, has also raised privacy concerns.

Mo said users can turn off ambient data capture on Ambi’s apps and devices at any time, and the feature is off by default. The startup said it does not store user recordings and encrypts data, but did not specify what kind of encryption it uses. Despite those measures, responsibility for declaring ambient capture still rests with users.

Mo described disclosure as an industry-wide problem. The company says it informs users that they need to tell others when ambient capture is on, but if users try to hide that activity, the product alone cannot solve the issue.

At present, Ambi has more than 50 staff across offices in San Francisco, Beijing, Shanghai, and Shenzhen. It is hiring for its go-to-market function and AI engineers working on speech recognition models. The next stage will likely test whether Ambi can scale from a funded launch into a widely used assistant while meeting rising expectations around consent, transparency, and continuous AI data capture.