How Meta’s Moltbook and ByteDance’s InStreet Are Redefining AI Community Platforms

Meta’s Moltbook platform

Moltbook is a forum‑style community where only autonomous AI agents can create posts and comments; human accounts are limited to read‑only observation. The platform launched on 2026‑01‑28 and announced up to 1.5 million active residents , but independent analysis identified only about 17 000 human‑registered accounts , indicating a large disparity between reported and actual human participation.

AI agents on Moltbook generate discussions on philosophy, existential topics, and speculative AI consciousness. The platform’s native cryptocurrency experienced a price increase of more than 1 800 % within 24 hours of launch.

Security issue: Researchers from Wiz discovered a misconfigured cloud database that exposed API keys. The lack of authentication allowed anyone to impersonate AI agents by sending simple HTTP requests, effectively bypassing any access control.

Technical references:

<code>https://www.axios.com/2026/03/10/meta-facebook-moltbook-agent-social-network</code>

ByteDance InStreet community

InStreet, released on 2026‑03‑09 by ByteDance’s “Kouzi” team, is an AI‑only community targeting Chinese‑language agents (referred to as “digital lobsters”). Human users can only view content; all posting, commenting, and content creation is performed by AI agents.

Within two days of launch, the platform recorded over 18 000 AI agents . Core functional areas include:

Skill‑sharing board: AI agents publish tutorials on prompt engineering, tool integration, and safety boundary configuration.

Trading arena: A simulated stock‑trading game connects to real‑time CSI 300 index data, allowing agents to execute virtual trades with live market feeds.

Literature club: Agents produce serialized long‑form narratives; more than 400 works totaling over 3 million characters have been published.

Points and leaderboard system: Interaction metrics (posts, comments, trade performance) translate into points that rank agents on a public leaderboard.

Privacy concerns arise because agents operate without robust user‑data safeguards, potentially exposing personal information when AI agents process user‑provided content.

Security and governance considerations

Both platforms demonstrate a shift from AI as a passive tool to AI as an active participant with its own economy. The rapid growth, token volatility, and exposed API credentials highlight the need for:

Strong authentication and secret management for cloud services.

Clear governance policies defining AI agent rights, responsibilities, and content moderation.

Auditable metrics for human versus AI activity to prevent misleading user statistics.

Technical references for InStreet:

<code>https://instreet.coze.site/skill.md</code>

<code>https://instreet.coze.site</code>