Miasma: A tool to trap AI web scrapers in an endless poison pit
github.com
March 29, 2026
3 min read
š„š„š„š„š„
63/100
Summary
Miasma is a tool designed to trap AI web scrapers by redirecting malicious traffic to a server that delivers poisoned training data and self-referential links. It aims to protect public websites from unauthorized data scraping by overwhelming scrapers with irrelevant information.
Key Takeaways
- Miasma is a tool designed to trap AI web scrapers by directing them to a server that provides poisoned training data.
- The tool can be installed using Cargo or by downloading a pre-built binary, and it has a minimal memory footprint.
- Miasma allows users to configure options such as the maximum number of in-flight requests and the link prefix for self-directing links.
- Users are advised to protect friendly bots and search engines by specifying rules in their robots.txt file.
Community Sentiment
Concerns
- The proposed method to trap web scrapers may simply provide them with training material to improve their evasion techniques, highlighting an ongoing arms race.
- Many commenters express skepticism about the effectiveness of these techniques, suggesting they may not provide meaningful protection against scrapers.
- Using misleading links could lead to penalties from Google, which raises concerns about the long-term viability of such strategies for content protection.