OpenClaw security fears lead Meta, other AI firms to restrict its use

In a recent communication, Jason Grad, CEO of a tech startup, alerted his team of 20 employees about the potential risks associated with the AI tool OpenClaw, previously known as Clawdbot. His late-night message, punctuated with a red siren emoji, warned that while the tool has gained attention on platforms like X and LinkedIn, it remains unvetted and poses significant risks to their work environment. Grad emphasized the importance of keeping OpenClaw off company devices and work-related accounts. This sentiment is echoed across the tech industry, as executives at major firms like Meta have also expressed serious concerns over the use of OpenClaw. One anonymous Meta executive mentioned that he has advised his team to avoid using the software on their work laptops, warning that failure to comply could lead to job loss. He described OpenClaw as unpredictable, highlighting the potential for privacy breaches in secure settings. OpenClaw, created by solo developer Peter Steinberger, was launched in November as a free and open-source tool. Its popularity has surged recently, thanks to contributions from other developers who have added features and shared their experiences on social media. In a significant development, Steinberger has joined forces with ChatGPT developer OpenAI, which plans to maintain OpenClaw's open-source status and support its development through a dedicated foundation. Setting up OpenClaw requires basic software engineering skills, after which it can autonomously manage tasks on a user's computer, such as organizing files, conducting online research, and facilitating e-commerce activities. In light of its capabilities, cybersecurity experts are urging companies to implement strict guidelines for its use, reflecting a growing trend where organizations prioritize security over the allure of experimenting with new AI technologies.