Recently, the open-source AI tool OpenClaw (often referred to as "Longxia" or "Lobster") has attracted considerable attention due to its autonomous capabilities in executing computer tasks. These tasks span from duplicating research reports automatically to continuously monitoring markets around the clock, demonstrating its robust functionality. Nevertheless, not all sectors are welcoming this innovation with open arms.

Beginning on March 11, numerous securities firms have frequently issued internal compliance alerts, explicitly banning the unauthorized installation, deployment, and utilization of OpenClaw on corporate networks and information systems. Should there be a necessity to use it, prior approval and reporting are mandatory. Certain securities firms have gone a step further, demanding the immediate removal of already installed instances and implementing safeguards for personal computers accessing the company network, with non-compliance leading to accountability measures.

Banks, too, are generally adopting a cautious stance, as OpenClaw’s extensive system operation permissions and remote control features conflict with the stringent security principle of segregating internal and external bank networks. The transparency of its source code and the internal network security threats posed by remote control functionalities have discouraged banks from adopting it.

Despite OpenClaw’s widespread popularity within the developer community, financial sectors such as securities firms and banks maintain a high level of vigilance regarding its security implications. Consequently, the industry is progressively reinforcing security boundaries for AI tools.