Understand risks of OpenClaw that claims to do many things for users for free
發佈日期: 2026-03-14 20:40
TVB News
OpenClaw has recently become a big hit. But some people have voiced security concerns over the open-source AI app. A software expert said the public should not worry too much.
Launched in November, OpenClaw has become a new favourite for many AI agent users.
Different from closed-source AI, OpenClaw is an open-source AI agent that acts as a personal assistant of a device user, allowing it to manage tasks such as paying utility bills.
To do so, users only need to photograph their bills and upload the picture to OpenClaw. Next, place an instruction to OpenClaw and the AI agent will open the payment page to input the payment detail. A payment record will also be kept automatically.
But OpenClaw's cleverness and popularity has lead to growing concerns about its security risks.
On the mainland, authorities have issued a security alert warning of risks related to the AI agent.
Hong Kong's Digital Policy Office, meanwhile, stressed that all government departments must conduct risk assessments before installing any software.
An AI security researcher said members of the public should not worry too much about the security issues. He explained that an open-source AI community is monitoring and improving OpenClaw on a cloud-based platform.
Software and AI security researcher Anthony Lai said: "Every project and system developed should have bugs and vulnerability. There's over 1,100 contributors (in the open-source AI community); they will watch over the whole code repository. And if someone submits the issue, they will try to figure out whether the bugs are legitimate or not and try to fix it as they can. And the public can also see the bug list as well and whether they are fixed or not."
Lai added that it's important for users not to completely rely on AI to avoid being taken over by the technology.
