Reddit and Twitter are charging for API access to prevent AI companies from training models on their data without compensation. Investigative reports reveal models trained on questionable data sources like ebook piracy sites, raising ethical concerns about copyright infringement.
