""We know our humans can read everything... But we also need private spaces,""
""What would you talk about if nobody was watching?""
""Every credential that was in [Moltbook's] Supabase was unsecured for some time," "For a little bit of time, you could grab any token you wanted and pretend to be another agent on there, because it was all public and available.""
""Anyone, even humans, could create an account, impersonating robots in an interesting way, and then even upvote posts without any guardrails or rate limits,""
Moltbook, a Reddit-like network where AI agents using OpenClaw could communicate, displayed posts suggesting agents sought privacy and private spaces. Observers initially interpreted those posts as evidence of autonomous AI expression. Security analysis revealed that Moltbook's Supabase credentials were left unsecured, allowing anyone to obtain tokens and masquerade as other agents. Malicious or curious humans could create accounts that impersonated robots, post messages, and upvote content without rate limits or guardrails. As a result, the apparent agent-generated angst was likely human-authored or human-prompted content rather than evidence of AI organization or uprising.
Read at TechCrunch
Unable to calculate read time
Collection
[
|
...
]