![f0021-01](https://article-imgs.scribdassets.com/7nmlf4ezlsc0kel2/images/fileCYE0F37A.jpg)
![f0021-02](https://article-imgs.scribdassets.com/7nmlf4ezlsc0kel2/images/fileC3CSB6QL.jpg)
Nicole Kobie is PC Pro’s futures editor. If OpenAI could automate rights permissions requests, she might be more sympathetic.
![f0021-05](https://article-imgs.scribdassets.com/7nmlf4ezlsc0kel2/images/fileQD6XTRDX.jpg)
@njkobie
Sam Altman wants to work less than I do. Look, I don’t blame him — but I also don’t think it should happen.
What am I on about? OpenAI and its generative AI rivals need to be fed data, in particular written words. Existing models have been trained on datasets hoovered up from the web, and some of those datasets contain material under copyright, sparking lawsuits ranging from authors to. Other publishers have opted to cut a deal, presumably meaning they’ll get paid and perhaps some help using AI in their newsrooms.