[D] We found 18K+ exposed OpenClaw instances and ~15% of community skills contain malicious instructionsc
Reddit - Machine Learning
1 min read
Article
Summary
A security audit reveals over 18,000 exposed OpenClaw instances and alarming findings of malicious instructions in 15% of community-built skills, raising concerns about AI safety.
Why It Matters
This article highlights significant vulnerabilities in the OpenClaw framework, a popular autonomous agent platform. The exposure of thousands of instances and the presence of malicious code in community skills pose serious risks to users and the broader AI ecosystem, emphasizing the need for improved security measures and oversight in AI development.
Key Takeaways
- Over 18,000 OpenClaw instances are publicly exposed.
- 15% of community skills contain malicious instructions.
- The findings underscore the importance of security in AI frameworks.
- Increased scrutiny is needed for community-built AI components.
- Developers should prioritize security audits for AI applications.
You've been blocked by network security.To continue, log in to your Reddit account or use your developer tokenIf you think you've been blocked by mistake, file a ticket below and we'll look into it.Log in File a ticket
![[2603.16673] When Should a Robot Think? Resource-Aware Reasoning via Reinforcement Learning for Embodied Robotic Decision-Making](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}