• michaelmrose@lemmy.world
    link
    fedilink
    English
    arrow-up
    15
    ·
    16 hours ago

    Notably the risks aren’t simply that this will identify the undocumented its that it will provide a pretext to disappear almost anyone who isn’t white.

    The question is does any given non-white person look enough like one of over 10M people to get falsely flagged given agents a pretext to remove almost anyone they stop. Facial recognition of one person against a large enough database will almost always provide at least on possible match this is especially true if it matches against possible aliens and not against possibly aliens and citizens. EG a query of both will likely turn up Betty Sue Smith is herself and maybe a possible match with Known Deportable but a search against only aliens may return only Known Deportable with agents dismissing any claim by Betty that her ID must be fake.

    • Almacca@aussie.zone
      link
      fedilink
      English
      arrow-up
      5
      ·
      15 hours ago

      it will provide a pretext to disappear almost anyone who isn’t white.

      Pretty sure you don’t need a.i. glasses to figure out what colour someone is.

      • Canigou@jlai.lu
        link
        fedilink
        English
        arrow-up
        6
        ·
        11 hours ago

        No but you can say “The AI told me to detain them” and not take any responsibility. “I was just answering orders” is sadly too common of an excuse in dooing the worst things…

        • Almacca@aussie.zone
          link
          fedilink
          English
          arrow-up
          1
          ·
          10 hours ago

          And if anyone believes that absolves them of responsibility, they deserve all the police stating they get.