Levels launches Blocked for AI Reply
Pieter Levels launched Blocked for AI Reply, a searchable database and API tracking over 3,700 X accounts deploying automated LLM engagement spam. The project includes evidence logs and copy-paste prompts enabling browser agents to mass-block flagged bots.
Cheap LLM inference has turned social media replies into an algorithmic engagement-farming ground that outpaces platform moderation. Deploying browser agents to automate mass-blocking turns agentic AI into a defensive shield, though manual oath-based appeals show how chaotic anti-spam governance remains.
- –Multi-source evidence tracking: Aggregates accounts flagged by Pieter Levels, Ryan Vogel's Clankwall detector, and Dillon Mulroy's list, storing full reply context and timestamps.
- –Agentic browser blocking: Offers a prompt for browser agents (Claude, Grok, ChatGPT Agent) to download usernames and automate blocking directly in the user's logged-in session.
- –Platform decay and bot economics: Highlights the growing problem of low-cost LLM bots farming impressions on viral posts to monetize social media accounts.
- –Community appeal model: Accounts can request removal by posting a pledge swearing off AI reply automation, attempting human review where algorithmic bans fail.
DISCOVERED
6h ago
2026-11-19
PUBLISHED
7h ago
2026-11-19
RELEVANCE
AUTHOR
levelsio