360 recently discovered three vulnerabilities in OpenClaw: one high-risk and two medium-risk.

360's vulnerability discovery AI recently uncovered and reported three high-value vulnerabilities in OpenClaw: one high-risk and two medium-risk. All newly discovered vulnerabilities have since been patched and publicly disclosed by 360. It is understood that these three newly discovered vulnerabilities directly target the core operating mechanism of the AI ​​agent, directly impacting the core security of user devices, data, and accounts. (The Paper)