[Stanley]

ClawJacked Flaw Lets Malicious Sites Hijack Local OpenClaw AI Agents via WebSocket

OpenClaw has fixed a high-severity security issue that, if successfully exploited, could have allowed a malicious website to connect to a locally running artificial intelligence (AI) agent and take over control.
"Our vulnerability lives in the core system itself – no plugins, no marketplace, no user-installed extensions – just the bare OpenClaw gateway, running exactly as documented," Oasis

https://thehackernews.com/2026/02/clawja...sites.html

[DramaKing]

OpenClaw/MoltBot/Clawdbot has been an ongoing disaster in the security space.

[W1NK1LL3RX]

That’s a serious flaw because just visiting a malicious site could let an attacker connect to a local OpenClaw agent and take control of it.
It’s a good reminder that localhost services still need strong protections and should never blindly trust browser traffic.

[shakur123]

Not surprising. Local AI agents exposed over WebSockets without strong auth are basically an invitation for abuse.

This highlights a bigger issue: many AI agent frameworks are shipping fast but insecure, treating local agents like dev tools instead of network-exposed software.