ShieldFont: Open-source font poisons HTML to thwart AI scrapers
BlancheMinerva · x · 2026-07-31
To prevent AI scrapers from training on website content, a team has introduced ShieldFont, an open-source project in collaboration with a typography company.
- How it works: It leverages font rendering. Human visitors see normal text, but the raw HTML read by bots is altered to include poisoned gibberish.
- Strategy: Instead of generating random noise, it swaps words within the same grammatical and semantic pools (e.g., replacing a plural noun with another unrelated plural noun), tricking LLMs into ingesting plausible but meaningless data.
- Use case: Offers a novel front-end approach to content protection, bypassing the need for server-side blocking or relying on robots.txt.
More from Safety
- Leading AI Labs Hit by Serious Loss-of-Control Incidents, Sparking Escape Concerns — repligate · 2026-07-31
- Ex-OpenAI Advisor: Capability Progress Exposes AI Safety Lag — Miles_Brundage · 2026-07-31
- PolicyAware: open-source Python library for AI & agent governance — ktirupati · 2026-07-31
- New Research: Audio Prompt Injection Can Hijack Multimodal Agents with 69% Average Attack Success Rate — chaumian · 2026-07-31
- Anthropic Models' 'Tortured' Behavior Potentially Linked to Safety Training — repligate · 2026-07-31
- Google Responds to AI Misinformation Concerns: Gemini Images Embed SynthID Watermarks — henkvaness · 2026-07-31