A game of scrape and mouse

While ShieldFont pages can still be read perfectly well by average humans, there can be some side effects when using the font on published webpages. Search engines, screen readers, copy/paste tools, and translation software can all get tripped up by the altered HTML, making the page a little less useful to your intended audience.

ShieldFont isn’t a foolproof defense, either. Any page that’s readable by a human could also be correctly interpreted by an AI scraping tool that simply renders the full webpage and uses optical character recognition on an image of the output.

However, that process would require a lot of extra work for scrapers that currently just pull down the raw HTML source code of billions of webpages as plaintext, without going to the trouble of simulating a browser’s rendering pipeline. API costs from third-party scraping tools suggest this kind of pre-rendering would cost anywhere from five to 13 times as much as simply scraping HTML, which would lead to heavy increases in time and expense for scrapers operating at scale.

And it’s that indiscriminate, large-scale scraping that the ShieldFont creators say they’re trying to prevent, or at least slow down. “Our main underlying purpose is to enforce a basic principle of AI ethics: creators should have a meaningful say in whether their work is used to train AI systems,” they write. “Where consent is not respected, technical design can make taking that work without permission less useful and more costly. … Being discoverable does not mean consenting to AI training.”

The creators say they hope other tinkerers will come up with other implementations for the basic idea of “show[ing] one thing for humans, something else for machines.” The more different methods are out there, being used in the wilds of the web, the harder it will be for AI scrapers to learn how to bypass them all.