3

Inappropriate / outdated / harmful search results may be required to remove upon request of the concerned individual. Google etc. have ways how to request removal.
Kagi doesn't yet.

Give "feedback" to a specific search result which triggers a request to Kagi system to investigate the legitimacy of the request for removal and then Kagi acts upon that to reply to the user and remove or not remove.

  • Vlad replied to this.

    Rene_b Can you give an example of what criteria would be used to decide whether content is appropriate? I am assuming you are referring to content that would be legal to begin with?

      Vlad

      I am mostly referring to GDPR-related and "right to be forgotten" or "defamatory" mentions.

      Where persons deal with an issue where your name is e.g. in an online article and you'd NOT WANT it there because it damages your professional reputation.
      In my recent case, I had succeeded in requesting removal of my personal information at the website operator side AS WELL AS succeeded with Google to assist in removing specific (defamatory, damaging) search requests for "my name".

      For Kagi Search, I'd want to at propose to have to option to ask for "re-crawl", have an option to see "where the result is from" or even request manual intervention to have a person at Kagi decide "blocking" a specific search result for a given search term.

      • Vlad replied to this.

        This will not be as easy as it sounds, unless Kagi hires a lot of lawyers to determine each case. When is information on a person damaging to find online and when is it necessary? Some cases are very clear, others are not. Criminals can and are using Google takedowns to scrub search results from their crimes, with the "right to be forgotten" protocol. So it cannot be as easy as letting any person mentioned in an online article get it removed because she doesn't like what is written about her.

        Then we have things like doxxing, which is of course something very nasty, when people's place of living are published. But sometimes they have published it themselves, etc.

          Rene_b We have an indicator if the content comes from our own index (flying dog shows in domain info). If it comes from an external index, it should reflect as soon as it is reflected there (not that we use multiple indexes so you'd need to request removal from all of them). I wish there was a more centralized way to do this.

            No one is typing