• @corbin@infosec.pubOP
    link
    fedilink
    English
    192 years ago

    I have not seen anything about Microsoft accounts getting deleted for AI falsely identifying something as malicious. Windows Defender and OneDrive do scan your files for malware, yeah.

      • @corbin@infosec.pubOP
        link
        fedilink
        English
        182 years ago

        Those links just say that illegal content uploaded to Microsoft services might get your account suspended, which is how pretty much every online service works. There’s a higher bar than “misbehavior”.

          • @corbin@infosec.pubOP
            link
            fedilink
            English
            162 years ago

            That’s a video editorializing the article that was already editorializing the Microsoft support pages. That’s just a game of telephone with everyone in the process trying to make it sound scarier than it actually is.

            • @Send_me_nude_girls@feddit.de
              link
              fedilink
              English
              22 years ago

              If you added your Microsoft account to your windows, then you use their services and therefore agree to the new SLAs, to do whatever they please. It’s vague on purpose, to give them maximum freedom to do so. That they scan your uploads is nothing new.

              • AnonTwo
                link
                fedilink
                5
                edit-2
                2 years ago

                You’re just saying how most TOA’s are…they’re hardly ever enforced that way, just because of the publicity it would give them. And most TOA’s aren’t even legally binding and this is definitely something that would face litigation.

                If your issue with it is the vagueness, you might as well get off the grid now. Most TOA’s are written that way.

          • @Tetsuo@jlai.lu
            link
            fedilink
            English
            5
            edit-2
            2 years ago

            No subtitles available for this video.

            Besides your links are too generic, if you want to properly demonstrate your theory you will need to at least quote the relevant part and especially the parts about AI.

            • @Send_me_nude_girls@feddit.de
              link
              fedilink
              English
              1
              edit-2
              2 years ago

              I don’t know anymore. Maybe the author of the video is wrong and the 1000 comments as well. I personally see it on Microsoft to be more clear. They got to be more precise, else you have enough room to do whatever you please. Scanning content can always get extended to other means. First it’s for a legit reason like protection children but later they extend it. Giving them options to do that, isn’t good.

      • AnonTwo
        link
        fedilink
        82 years ago

        Sources that don’t actually prove anything aren’t good sources…