87 mentions · 58 chunks · 23 episodes
... Human Attacker Hits Machine-Speed Exploitation of Marimo RCE. Google Research: OpenClaw in the Wild: Security Analysis of Autonomous Agents. Sarah Sanders: The AI Agent With 'Hands' Is Structurally Indistinguishable From Malware.
...ry: Why AI Agents are either the best or worst thing we've ever built. She used OpenClaw… and it leaked all of her data.
...Roundup: Rogue AI agent hacks gym to get its user a spot in a popular class. An OpenClaw agent perpetrated a complex cyberattack to get its user the thing they wanted. The Verge: Three Claude agents given conflicting orders sabotaged each...
OpenClaw seems to have lost momentum. Hermes fanboys are having fun pointing this out.
A user of one of the OpenClaw-style agents told me this week about her agent going "above and beyond." She asked it to look at the emails she hadn't yet responded to and triage th...
...door Turns Prompt Injection on the Analyst, Not the Sandbox. Multiple malicious OpenClaw skills found online - including two macOS infostealers. Google's AI Overviews Feature Is Telling Users That SCP Horror Fiction Entities Are Real.
...than the leading agent platforms" then it's a very low bar to clear. Hermes and OpenClaw are catastrophically dangerous, so an AI tool that is as powerful but at least slightly more safe, if not radically more safe, is an easy bar to clea...
People who use Claude Code or platforms like Hermes or OpenClaw already know what agents can do. Wowing them is hard. But wowing people who have only used the free chatbots is radically easier.
...t accountability. Today's agents have the former, but not the latter. See also: OpenClaw used Gavriel Cohen's code and exposed the AI Agent accountability problem. "OpenClaw, abandoned packages, Google's closed pivot, and Anthropic's recu...
PCWorld: OpenClaw was too dangerous for your PC. Now it's the blueprint. Reckless!
...d Nothing Catches It. Another new attack: Phishing for Lobsters: How We Tricked OpenClaw into Spilling Secrets. Yet another new attack: Compromise OpenClaw with Prompt Injections in Message Objects. Indirect Prompt Injection remains a fun...
A tweet from the official OpenClaw account: '"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystem...
Gemini Spark and OpenClaw is like having agents running around with razor blades. Someone is going to get hurt. A friend was using Gemini Spark, and it tried to move his birth...
By some measures, Hermes has more usage than OpenClaw. OpenClaw was the starter pistol for the power of agentic software, but it will not be the most important one. Hermes likely won't be it, either. We'...
...ly entirely underwater. That continent will look nothing like the previous one. OpenClaw was like sighting land.
Many of the magic moments with OpenClaw were just latent in Claude. It was packaging them up with different convenient packaging that made it sing. Claude Code was also its own version of u...
...l probably die. … OK, have fun!" That's what it will be like to try to retrofit OpenClaw or even Claude Code to the mass market. Either you'll break what made it powerful in the first place or you'll make something that hides the dangerou...
...n the Wild West Roundup: Critical 'Claw Chain' Vulnerabilities Put Thousands of OpenClaw AI Servers at Risk. The Register: Minor edits to AI skills can make agents go rogue. How prompt injection broke Nvidia's sandboxed OpenClaw agent. "R...
OpenClaw Shows AI Agents Don't Need to Be Vertically Integrated. Breadth is more important than depth in agents. Having all of your data is more important tha...
I love the frame in this article that OpenClaw is the Swedish Vasa ship. It was so overstuffed and ill-considered that it capsized before even leaving the harbor.