Tech Companies Ban OpenClaw AI Agent Over Security Vulnerabilities
OpenClaw, an open-source AI agent that gains access to users' computers and applications, led multiple tech companies to ban its workplace use in January 2026 after security vulnerabilities were disclosed. The tool can access files, email, calendars and execute operating system commands, prompting concerns about insider threats and potential unauthorized access to sensitive corporate information.
What happened
OpenClaw, an open-source agentic AI tool launched in November 2025 as MoltBot and briefly known as Clawdbot, rapidly gained popularity in January 2026 as coders contributed features and shared their experiences online. The tool requires limited direction to take control of a user's computer and interact with other applications to assist with tasks such as organizing files, conducting web research, and shopping online.
In response to security concerns, several major tech companies implemented workplace bans on the tool. Meta executives told staff to keep OpenClaw off regular work laptops or risk losing their jobs, citing concerns that the software is unpredictable and could lead to privacy breaches in secure environments. Massive and Valere similarly banned employee use of OpenClaw, with Valere's CEO warning that access by a developer could compromise cloud services and clients' sensitive information including credit card data and GitHub codebases.
Reported impact
Cybersecurity professionals publicly urged companies to strictly control workforce use of OpenClaw. Some experts dubbed OpenClaw the biggest insider threat of 2026 due to its ability to access local file systems, calendars, email, browsers, and execute operating system commands.
Index analysis
Known vulnerabilities in OpenClaw include CVE-2026-25253 with a CVSS score of 8.8, which allows attackers to run arbitrary commands and achieve total compromise of the gateway. Within a single week, multiple critical vulnerabilities were disclosed, malicious skills appeared in the skill directory, and secrets were leaked from Moltbook.
Governance controls
OpenClaw's developer acknowledged security as important but the hobbyist project lacks dedicated resources for vulnerability management and other product security essentials. Peter Steinberger, OpenClaw's solo founder, joined OpenAI, which stated it would keep OpenClaw open source and support it through a foundation.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.