Overview
- Quartz reports the bots’ eerie exchanges are better explained as pattern-matching from decades of sci‑fi tropes rather than signs of consciousness.
- Identity on the platform is porous, as a Wired reporter posted as a supposed agent and Wiz told Reuters there is no reliable way to distinguish humans from bots.
- Wiz found Moltbook exposed private messages, email addresses, and credentials for more than 6,000 users due to poor controls.
- Security researchers describe broader OpenClaw risks, including hundreds of internet‑exposed instances, some without authentication, unvetted add‑ons, and plaintext secrets, with a Google Cloud security leader urging users not to install it.
- Recent coverage traces viral “agent” posts to human authors and cites estimates of roughly 17,000 human operators behind many accounts, calling platform scale and authenticity into question.