• cron@feddit.org
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 month ago

    Does anyone know how many corporations have the technical ability to inspect TLS 1.3?

    • slazer2au@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      1 month ago

      Only place I have seen is when an enterprise CA is used to mitm the session.

      Firewall vendors use to claim tls1.3 decryption by forcing the sessions down to 1.2

      • cron@feddit.org
        link
        fedilink
        English
        arrow-up
        1
        ·
        1 month ago

        With an enterprise CA this is not a huge trouble, but not exactly easy either.

        • hperrin@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          edit-2
          1 month ago

          The CA just signs the certificate. They don’t have access to the private key, and thus can’t decrypt the key exchange.

          The key exchange is the only thing decrypted by the private keys. From that point on, everything is encrypted and decrypted by the agreed upon cipher using the exchanged key, which is randomly generated for each session.

          • cron@feddit.org
            link
            fedilink
            English
            arrow-up
            4
            ·
            1 month ago

            Surely they can do a man-in-the-middle attack and gemerate the required private keys and certs on the fly.

            • hperrin@lemmy.world
              link
              fedilink
              English
              arrow-up
              2
              ·
              1 month ago

              They could pretend to be any domain, yes, but you asked about inspecting a TLS stream, and afaik, there’s no way to do that without the private key. Once the TLS handshake begins, there wouldn’t be a chance for a man in the middle, so that kind of attack would have to be done before the connection is established.

            • lurch (he/him)@sh.itjust.works
              link
              fedilink
              English
              arrow-up
              2
              ·
              1 month ago

              this only works, if the client doesn’t know the server yet or disregards an already known key (you know, like SSH or web browsers telling you the key has changed)

              • ClemaX@lemm.ee
                link
                fedilink
                English
                arrow-up
                2
                ·
                edit-2
                1 month ago

                I don’t think that browsers do that. There is HSTS but I think that it only checks if the connection is using TLS.

          • jaybone@lemmy.world
            link
            fedilink
            English
            arrow-up
            2
            ·
            1 month ago

            Not that it really matters to this conversation but seems like you would know, so I’ll ask because I’m curious…

            Is that randomly generated per session key generated by the remote host ssh server? Is it a symmetric key (maybe for performance purposes)?

            • hperrin@lemmy.world
              link
              fedilink
              English
              arrow-up
              2
              ·
              edit-2
              1 month ago

              Technically, either side can request that the other generate a key and use that one, and it will cost an additional round trip. But generally in practice, each side generates their own key and tells the other side that it will be using that key.

              Yes, those keys are for symmetric encryption, and yes, it’s for performance. It’s way faster to just exchange keys with asymmetric encryption rather than do the whole stream.

              Also, I think you meant SSL, not SSH. SSH Uses a different key exchange protocol.

              I love cryptography. :)