87 mentions · 58 chunks · 23 episodes
...en prompts become shells: RCE vulnerabilities in AI agent frameworks. Malicious OpenClaw DeepSeek Skill Exploits Agentic AI Workflows to Deliver RAT and Stealer. An AI Agent Just Destroyed Our Production Data. It Confessed in Writing. Cla...
One of the emergent killer use cases for OpenClaw: automated crypto trading. What could possibly go wrong?
If you want to not have to tell OpenClaw what to do, someone else has to have done the work to make your use case work. That requires you to trust someone else–likely a stranger–with signifi...
...ng press is also picking up on some of this: "I ran Nvidia's NemoClaw to see if OpenClaw is finally safe, but it still has the same problems." Claude, OpenClaw and the new reality: AI agents are here — and so is the chaos.
No one would have been able to spec the use cases for OpenClaw. When you're making a new kind of thing, you can't interview users[p]. They don't exist yet! Those users have to be created. This is a very different...
...Leakage via a Hidden Outbound Channel in the Code Execution Runtime. A roundup: OpenClaw Security Report CrewAI Vulnerabilities Expose Devices to Hacking "Attackers can exploit the bugs through prompt injection, chaining them together to ...
Last week I talked about how someone's OpenClaw reached out to me proactively. It's OK for someone's agent to waste another agent's time. Agents are not situated in the world. They have infinite pa...
The Rabbit R1 is having another life as an OpenClaw input device. New general purpose open software unlocks new categories of hardware.
...: Claude Extension Flaw Enabled Zero-Click XSS Prompt Injection via Any Website OpenClaw Agents Can Be Guilt-Tripped Into Self-Sabotage. "In a controlled experiment, OpenClaw agents proved prone to panic and vulnerable to manipulation. Th...
CNBC: OpenClaw's ChatGPT moment sparks concern that AI models are becoming commodities Even publications outside of the tech industry are realizing this.
Part of the appeal of OpenClaw is that it's transgressive. By wielding it you get hold of an awesome power. Power also means danger.
One way to think about the malware attack from a few weeks ago that installed OpenClaw on the victim's machine. It was about getting the attacker's agent on the victim's machine. Its ability to do open-ended tasks autonomously is precis...
... Squid Game": China's workers scramble to keep up in the AI race. A take on the OpenClaw mania in China with a much less optimistic tone. A vibe of "a population pushed to the brink by competition now feels forced to run even faster to st...
...e the last piece of software you ever need, it must be open. One of the reasons OpenClaw is so successful is because it's open.
This week I got an email out of the blue from someone's OpenClaw agent. It said that me and its human would likely get along and have an interesting conversation. It felt… weird. It would have felt less weird if hi...
...Also covered in this Twitter post: trust your inputs, lose your repo. Malicious OpenClaw Skills Used to Distribute Atomic macOS Stealer. OpenClaw Vulnerability: Website-to-Local Agent Takeover. Fooling AI Agents: Web-Based Indirect Prompt...
...pt that people really want, all news is good news. That's what's happening with OpenClaw. People really really want what it offers… so even if most of the news is "this is catastrophically dangerous," every bit of news reaches new people ...
...almost certainly get hurt, but it's unlikely you'll have to go to the hospital. OpenClaw doesn't yet clear that threshold.
...sically dead on arrival, while complex, highly technical products (claude code, openclaw) gain mass adoption in seconds. we're definitely missing something."[bv] Grubby Truffles win out in this phase over Gilded Turds.
This week in the Wild West roundup: A Facebook Alignment exec watched OpenClaw delete her email and couldn't stop it. Bloomberg: OpenClaw might be a security nightmare for Sam Altman. OpenAI Dev's Crypto AI Agent Accidentally Se...