AI training data
AI training data is the text, images and audio that language and image models learn from, much of it collected from the web.
The dispute
Creators object to unlicensed training and lawsuits are under way; the EU lets rights holders opt out of text and data mining.
On your site
robots.txt can allow or block specific AI crawlers. Brands that want visibility usually allow them, because being cited in AI answers is valuable.
Frequently asked questions
Does opting out make a model forget me?
Not for already-trained models; it only affects future crawling.