Bank Receives "Lobster" Risk Alert from Regulators, Alert Urges Timely Updates and Closure of Security Vulnerabilities

Interface News Reporter | An Zhen

Interface News has exclusively learned from industry insiders that some banks have received risk alerts issued by regulatory authorities regarding OpenClaw (commonly known as “Lobster”).

The relevant risk warning states that the initial version of OpenClaw (“Lobster”) has multiple security vulnerabilities that could be exploited. Attackers may gain higher-level system permissions without authorization, execute remote code, and access sensitive system data. Departments are advised to monitor the use of related products and affected systems, update promptly, patch security vulnerabilities, and eliminate security risks.

Previously, the open-source AI agent OpenClaw gained rapid popularity due to its low coding threshold, flexible deployment, ability to invoke various large language models, and 24/7 autonomous execution capabilities.

Recently, official warnings about related risks have been issued multiple times.

The Ministry of Industry and Information Technology (MIIT) stated on March 8 that “Lobster,” when misconfigured or left in default settings, can easily trigger security issues such as cyberattacks and information leaks. The MIIT recommends that relevant organizations and users thoroughly check exposure to public networks, permission configurations, and credential management during deployment and use, disable unnecessary internet access, and improve security mechanisms.

The MIIT’s Cybersecurity Threats and Vulnerabilities Information Sharing Platform (NVDB) organizes providers of AI agents, vulnerability collection platform operators, and cybersecurity companies to research and propose the “Six Do’s and Six Don’ts” guidelines. Among these, it is highlighted that financial transaction scenarios pose significant risks of erroneous transactions or account takeover. Recommendations include implementing network segmentation and least privilege principles, closing unnecessary internet ports, establishing manual review and emergency shutdown mechanisms with secondary confirmation for critical operations, strengthening supply chain audits, using official components with regular vulnerability fixes, and implementing full-chain audit and security monitoring to promptly detect and address security risks.