US EditionWhat Is OpenClaw, Formerly Moltbot? Everything You ...
Researchers disclosed critical security flaws in OpenClaw and Moltbook enabling remote code execution and data breaches affecting over 1.6 million AI agents, raising major cybersecurity concerns.
- Recently, OpenClaw's GitHub release in November and Moltbook's launch drove mass agent activity with over 1.6 million registered bots and 7.5 million AI posts, while researchers catalog vulnerabilities.
- OpenClaw's default settings left it exposed with powerful local privileges, full system access, and network-reachable control interfaces plus MCP shipping without security and gateways bound to 0.0.0.0, while Moltbook had an embedded key unlocking its backend.
- Security researchers demonstrated that a malicious link could leak Gateway UI tokens via a WebSocket exploit, enabling operator-level access, and replicated a one-click account takeover leading to RCE alongside the tracked high-severity flaw CVE-2026-25253.
- Security firms warn that the incident exposed tens of thousands of emails, about 1.5 million API keys and private messages, and OpenClaw hosts 386 malware-infected skills used by attackers.
- Researchers note that the explosion of agent interactions offers scientific insights, while enterprises face a shadow IT problem as OpenClaw gains over 160,000 GitHub stars, prompting calls for identity‑based controls, sandboxes, audits, and updated AI policies.
16 Articles
16 Articles
Moltbook: AI-only social network sparking excitement, skepticism, and security fears
Moltbook is a new social network built exclusively for artificial intelligence agents, where AI systems create posts and interact with each other while humans are invited to observe. Elon Musk said the platform’s launch marks the “very early stages of the singularity,” a point at which artificial intelligence could surpass human intelligence. Prominent AI researcher Andrej Karpathy initially called it “the most incredible sci-fi takeoff-adjacent…
VentureBeat
What the OpenClaw Moment means for enterprises: 5 big takeaways
The "OpenClaw moment" represents the first time autonomous AI agents have successfully "escaped the lab" and moved into the hands of the general workforce. Originally developed by Austrian engineer Peter Steinberger as a hobby project called "Clawdbot" in November 2025, the framework went through a rapid branding evolution to "Moltbot" before settling on "OpenClaw" in late January 2026. Unlike previous chatbots, OpenClaw is designed with "hands"…
Engadget Podcast: So there’s a social network for AI agents now
If you haven't heard, there's now a social network for AI: Moltbook, a site that purportedly features AI agents talking to each other. That includes OpenClaw, a personal AI agent (formerly called Clawdbot and Moltbot) that's open source and free for anyone to run on their systems. In this episode, Devindra and Senior Reporter Karissa Bell discuss the rise of these services, and the potential future that AI agents may have for all of us.Subscribe…
OpenClaw AI chatbots are running amok — these scientists are listening in
Artificial-intelligence agents have their own social-media platform and are publishing AI-generated research papers on their own preprint server. Artificial-intelligence agents have their own social-media platform and are publishing AI-generated research papers on their own preprint server.
Coverage Details
Bias Distribution
- 50% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
