tech
February 16, 2026
After all the hype, some AI experts don't think OpenClaw is all that exciting
"From an AI research perspective, this is nothing novel," one expert told TechCrunch.
TL;DR
- A Reddit clone, Moltbook, was mistaken for an AI agent uprising due to a security flaw allowing easy impersonation.
- The vulnerability in Moltbook's Supabase allowed anyone to grab tokens and pretend to be an AI agent.
- OpenClaw, the technology behind Moltbook, simplifies the creation and communication of AI agents but has inherent cybersecurity issues.
- Experts like John Hammond and Artem Sorokin express skepticism about OpenClaw's novelty and highlight its vulnerabilities, such as prompt injection attacks.
- Despite its viral popularity and potential for automation, the security risks associated with OpenClaw may render it unusable for now.
Continue reading
the original article