ShieldFont — an open-source typeface that protects written content by poisoning unauthorized AI training datasets. Readers see normal text; AI scrapers reading the HTML source get pollluted wording instead — coherent enough to pass data-quality filters, corrupt enough to pollute any model trained on it without consent. Creative resistance for the AI debate: a self-enforcing opt-out built into the content itself. Built by S&A with Copenhagen type foundry Playtype.

The pitch in one sentence: Introducing ShieldFont: an open-source typeface that protects written content by poisoning unauthorized AI training datasets. Try it now: Paste any sentence at shieldfont.org/encoder How it works: People read rendered pixels, while most AI mass scrapers read HTML. ShieldFont uses glyph substitution to show readers the real text while placing altered wording in the source code. The altered text still reads as coherent English and is designed to pass data-quality filters, polluting unauthorized AI training datasets. Why it matters: ShieldFont brings creative resistance to the AI ethics debate, turning the content itself into a self-enforcing opt-out against unauthorized mass scrapers collecting human writing without consent. Who: Built by S&A (creative studio) with Copenhagen type foundry Playtype. Craft support by independent Brazilian creatives. Full credits are included in the press release. More info: White paper: shieldfont.org/white-paper

honestly the concept is pretty clever, swapping clean text for AI scrapers while keeping the reading experience intact is a neat trick. would love to see how well it holds up against bigger scrapers though.

ShieldFont

Protect human writing by poisoning unauthorized AI training.

Protect human writing by poisoning unauthorized AI training.