• Allero@lemmy.today
      link
      fedilink
      English
      arrow-up
      6
      arrow-down
      1
      ·
      edit-2
      4 hours ago

      Is that conspiracy kicking in? Besides, it is still LLMs. Like, these are text-generating things, not weapon-grade AI. The absolute worst they can do is be used to assist hacking, and modern models coming from the US are already very good at vulnerability detection and pentesting. Not to mention the best defense against that is not AI, but standardized and strictly enforced network security.

        • Allero@lemmy.today
          link
          fedilink
          English
          arrow-up
          1
          ·
          4 hours ago

          Precisely. If we want to keep our systems safe, we need not an extra AI, but strict adherence to well-established practices. Vulnerability in networks comes from lackluster execution of well-known, well-documented security measures.

          Leave AI as an extra level of pentesting, it already does such things brilliantly.