Meta's Manus launches desktop app enabling AI agents to access local files and applications
Meta's recently acquired AI startup Manus launched a desktop application enabling its AI agent to directly access local files, tools, and applications on personal computers through a 'My Computer' feature. Previously cloud-only, the move positions Manus to compete with OpenClaw, the open-source AI agent that sparked recent industry momentum. Unlike OpenClaw's free, MIT-licensed offering, Manus operates as a paid subscription service.
Meta's Manus Launches Desktop App to Bring AI Agents Onto Local Devices
Meta's AI startup Manus, acquired in December 2025, launched a desktop application Monday that brings its general-purpose agent directly onto personal computers. The new 'My Computer' feature marks a significant shift from Manus's previous cloud-only architecture, where agents operated exclusively through web interfaces.
Key Capabilities
Through the desktop app, Manus agents can now:
- Read, analyze, and edit files on local drives
- Launch and control applications installed on the machine
- Manage complex multi-step tasks locally
- Create applications within minutes
- Organize large datasets (e.g., thousands of images)
The feature complements existing Manus integrations with Google Calendar, Gmail, and third-party platforms.
Competitive Positioning Against OpenClaw
The launch directly addresses OpenClaw's momentum in the AI agent space. OpenClaw, an open-source agent founded by Austrian developer Peter Steinberger in late 2025, gained significant traction with its local-device installation model. Nvidia CEO Jensen Huang called OpenClaw the "next ChatGPT" during a Tuesday appearance on CNBC's "Mad Money."
The critical difference: OpenClaw is free and distributed under an MIT license, while Manus is primarily a paid subscription service. OpenAI has also hired Steinberger, positioning OpenClaw as a competitive threat to Meta's AI agent ambitions.
Security and Control Measures
Manus addresses privacy concerns raised by security experts about granting AI agents local device access. The company implements explicit approval requirements before task execution, offering two control modes:
- Allow Once: Individual review for each action
- Always Allow: Pre-approved recurring actions for trusted operations
Meta's Broader AI Strategy
Meta's $2 billion acquisition of Manus, announced December 29, 2025, aims to integrate autonomous agent technology across Meta's platforms, including the Meta AI assistant. However, Chinese officials are reportedly scrutinizing the deal for potential violations of technology controls.
Manus was founded in China before relocating its headquarters to Singapore.
What This Means
Meta is directly competing for the emerging AI agent market dominated by OpenClaw's open-source momentum. By moving agents from cloud to local devices, Manus gains feature parity with OpenClaw while maintaining a paid-subscription business model. The success of this strategy depends on whether users will pay for capabilities they can access free through OpenClaw, and whether Manus's permission-based controls adequately address legitimate security concerns about local AI agent access.
Related Articles
Meta's Muse AI Agent App Hits 730,000 Downloads, Overtakes ChatGPT on iOS Charts
Meta's Muse AI agent app overtook ChatGPT as the top free iOS app in the U.S., racking up 730,000 downloads in its first five days, according to Sensor Tower. The app, powered by Meta's Muse Spark model family, marks Zuckerberg's biggest push yet into AI agents.
Meta Brings Muse AI Agent to Mac, Enabling Actions Across Files, Mail and Calendar
Meta has launched a Mac desktop version of Muse, its AI assistant that can act on files, messages, calendar, notes and mail within native apps. The release follows Muse's mobile and web debut earlier this month, which topped U.S. App Store charts.
xAI's Grok 4.6 Launches on Amazon Bedrock With 500K Context and Cross-Region Inference
xAI's Grok 4.6 is now available on Amazon Bedrock via both bedrock-mantle and bedrock-runtime endpoints, adding Converse API support, cross-Region inference profiles, and Bedrock Guardrails. The model offers a 500K token context window and four reasoning effort levels, with input pricing starting at $2.00 per million tokens on the global inference profile.
New Siri AI in iOS 27 Frustrates Users With Bugs, Slow Responses, and Lock-Screen Failures
Siri AI, Apple's overhauled assistant rolling out in iOS 27, is generating user complaints over network errors, broken device search, unlock requirements, slow Apple Watch responses, and missing features from classic Siri. The assistant remains labeled beta with no disclosed fix roadmap beyond new language support in October.
Comments
Loading...