Thanks a lot for this critical but qualified response! As you found out already, the new permissions are needed so that Copyfish can communicate with the UI.Vision XModule, which it uses for desktop screenshots, for desktop OCR. Of course, you only need to install the desktop OCR module if you want to OCR outside of the web browser (e. g. from desktop apps, desktop dialogs, images, invoices, PDF,...)

For more details, please see https://ocr.space/copyfish#desktop