OpenClaw’s New Mobile App Turns Its Open‑Source AI Agent Into a Pocket Assistant

The open‑source AI agent project OpenClaw has released its first official mobile app for iOS and Android, giving users a way to talk to their personal automation agent from their phone. According to press reports, the app lets you chat with the AI via text or voice, review proposed actions, and approve or reject tasks directly on mobile. Instead of running the full agent stack on the handset, the app connects to an OpenClaw Gateway on macOS, Linux, or Windows (via WSL2). Heavy model workloads remain on the desktop or server, while the phone acts as a secure remote control.

On mobile, OpenClaw supports voice conversations, content sharing, and optional access to the camera, location, photos, contacts, calendar, and reminders, so agents can trigger workflows tied to your real‑world context. At the same time, the developers emphasize security: they explicitly warn about prompt‑injection attacks and recommend strict authentication, clear policies for which tools the agent may call, process isolation, and explicit user approval before executing tasks. In a landscape dominated by browser extensions and cloud‑only copilots, OpenClaw’s combination of an open‑source core, local gateway, and native apps is likely to appeal to developers and power users who want both convenience and control.