• SayCyberOnceMore@feddit.uk
    link
    fedilink
    arrow-up
    1
    ·
    1 hour ago

    By measuring the manifestations, attackers can decrypt encrypted traffic

    How ?

    Even the research paper doesn’t appear to mention this (unless I missed something)

  • lurch (he/him)@sh.itjust.works
    link
    fedilink
    arrow-up
    1
    ·
    22 hours ago

    this is complete bullshit and relies on a well trained AI, but users, websites and browser setups are too different for it to reliably work in the wild.

    this is the whole fingerprinting overreaction all over again.

    • infiniteCAD@lemmy.blahaj.zone
      link
      fedilink
      arrow-up
      1
      ·
      2 hours ago

      yup, aint no way anyone’s training data on latency of every website and app ever. nor would that be possible bc of just so so many possible variations in the stack of execuation. this attack, if feasible for anything, would only work if the user’s stack is well known already. even then there’s way better ways to attack