🐯Stalecollected in 10m

OpenClaw Blocked by China's App Walls

OpenClaw Blocked by China's App Walls
PostLinkedIn
🐯Read original on 虎嗅

💡OpenClaw's China barriers expose real-world limits for AI agents in silos

⚡ 30-Second TL;DR

What Changed

Bypasses APIs using vision-based screen reading and human-like input simulation.

Why It Matters

Severely limits OpenClaw's mass adoption in China, confining it to niche geek use while giants reinforce closed ecosystems, potentially stunting global agent interoperability.

What To Do Next

Test OpenClaw on a PC VM with throwaway accounts to prototype desktop automations safely.

Who should care:Developers & AI Engineers

🧠 Deep Insight

Web-grounded analysis with 6 cited sources.

🔑 Enhanced Key Takeaways

  • South Korea's major firms Kakao, Naver, and Karrot Market have restricted or blocked OpenClaw on corporate networks citing unmanageable risks to information assets.[1][2]
  • Security analyses by SlowMist, Koi Security, and Palo Alto Networks identified hundreds of compromised ClawHub extensions with infostealers and warned of private data access combined with untrusted content exposure.[1]
  • OpenClaw's viral growth reached over 250,000 GitHub stars in under 30 days by early 2026, surpassing Linux and React repositories.[5]

🛠️ Technical Deep Dive

  • OpenClaw automates workflows by running directly on users' operating systems, enabling web browsing, file editing, system command execution, and interactions via modular 'skills' extensions hosted on ClawHub.[1][2]
  • Requires high-permission credentials like usernames, passwords, or cryptographic keys to perform tasks, amplifying identity security risks even when properly configured.[3]
  • Developed rapidly with 600 commits per day, much of the codebase AI-generated and refactored, running with full system permissions by default.[5]

🔮 Future ImplicationsAI analysis grounded in cited sources

OpenClaw restrictions will expand beyond China and South Korea to more enterprises globally.
Enterprise blocks by Kakao, Naver, and Karrot, plus security firm warnings, signal rising governance challenges for autonomous AI agents with deep system access.[1][2]
Adoption of Chinese AI models like MiniMax will increase for OpenClaw to reduce costs.
Creator Peter Steinberger recommends MiniMax models, providing a cost advantage over US models amid geopolitical tensions.[4][6]

Timeline

2024-11
Anthropic introduces Model Context Protocol (MCP), later used by OpenClaw developers.
2025-11
OpenClaw launches as early-stage AI agent, previously named Clawdbot and Moltbot.
2026-01
OpenClaw explodes to over 250,000 GitHub stars in under 30 days.
2026-02
China's industry ministry warns of security risks from insecure OpenClaw deployments.
2026-02
South Korean firms Kakao, Naver, and Karrot restrict OpenClaw on corporate networks.
📰

Weekly AI Recap

Read this week's curated digest of top AI events →

👉Related Updates

AI-curated news aggregator. All content rights belong to original publishers.
Original source: 虎嗅