Open Claw AI Agent Flaws Could Enable Prompt Injection and Data Exfiltration
Summary: China's CNCERT warned that OpenClaw, an open-source self-hosted autonomous AI agent formerly known as Clawdbot and Moltbot, carries weak default security configurations combined with privileged system access that attackers could exploit to seize control of an endpoint, including through prompt injection. PromptArmor researchers previously showed that link previews in messaging apps like Telegram and Discord can be turned into a data exfiltration path when the agent is manipulated into building an attacker-controlled URL carrying sensitive data, and threat actors have also pushed malicious GitHub repositories posing as OpenClaw installers to deliver Atomic and Vidar Stealer and the GhostSocks proxy malware.
Key Takeaway: CNCERT warned that breaches in critical sectors such as finance and energy could leak core business data, trade secrets, and code repositories, or even paralyze entire business systems - and one malicious repository became the top-rated suggestion in Bing's AI search results for OpenClaw on Windows.
Read the original at The Hacker News ↗
- Topics
- AI Security
- Related
- SECURITY AFFAIRS AI-CYBERSECURITY NEWSLETTER ROUND 8 Cyber Risks · Nov 15, 2026Pentagon Renames CDAO ‘Super Intelligence Office’ Cyber Risks · Nov 15, 2026186 Fake AI Domains Redirect Users to Click Fix Lures, Browser Malware, and Search Hijackers Cyber Risks · Nov 12, 2026