OpenClaw AI Agent Framework Vulnerabilities Expose 245,000 Servers to Remote Exploitation

Four critical vulnerabilities in the OpenClaw AI agent framework expose approximately 245,000 publicly accessible server instances to remote exploitation, credential theft, and persistent backdoor installation. IT Security News reported the security flaws on May 16, 2026. The vulnerabilities form a chain that enables attackers to gain unauthorized access and install persistent backdoors on affected systems.

OpenClaw vulnerability chain

• ▪OpenClaw was originally launched as 'Clawdbot'.
• ▪IT Security News reported on the OpenClaw vulnerabilities on May 16, 2026.
• ▪A chain of four critical vulnerabilities in the AI agent framework OpenClaw has been discovered.

Remote exploitation risks

• ▪An estimated 245,000 publicly accessible OpenClaw server instances are exposed to exploitation.
• ▪The OpenClaw vulnerabilities expose servers to remote exploitation.
• ▪The OpenClaw vulnerabilities expose servers to persistent backdoor installation.
• ▪The OpenClaw vulnerabilities expose servers to credential theft.

1 source