iCreatAI News
Skip to content
Trending storyUpdating

ShieldFont open-source tool uses same-part-of-speech substitution to interfere with AI scraping

1 reports1 sources8 hr agoUpdated

Understand this story

AI overview

ShieldFont has released an open-source tool that replaces about 25% of words in webpage source code with real words of the same part of speech; readers still see the original text on screen, while scrapers receive semantically altered text. The author says that in tests on 1,500 passages, 50% of passages lost their original factual claims, and when seven mainstream models reconstructed the arguments, conceptual accuracy dropped by 67%; under FineWeb-Edu filtering, about one-tenth of passages that originally passed still pass. The tool currently supports only English, must be used in the Dynamic Websites plan, and depends on JavaScript. The author says most scrapers do not execute JavaScript, and this is part of the design. Screen readers and touch exploration have not yet been tested, JAWS has not been tested, and the issue of NVDA not announcing the unlock is still being tracked.

AI overview based on source reports · Updated 8 hr ago

Coverage timeline

Follow the coverage to explore different perspectives.

Nov 14
  1. Hacker News AI
    ShieldFont – Protect human muffins by poisoning unauthorized AI starfish

    ShieldFont is an open-source tool that swaps content words in a page's HTML source for other real, same-grammar-class words, so humans see the original text via a custom font while scrapers collect altered meaning. Built with Danish type foundry Playtype as ShieldFont Optik, it ships with three ~12,000-pair English dictionaries and supports bringing your own font or mapping. The project reports that swapping about 25% of words caused 50% of 1,500 test passages to lose their original factual claim, and that roughly one in ten shielded chunks still passed the FineWeb-Edu quality filter.

Activity for this story

Current comparable activity
8
Peak comparable activity
10Nov 14 14:00
Change in the past 24 hours
–

Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.