No More Alice to Bob: Reality-based Models for Message Encryption and Key Management
Terence Spies

Citation
Terence Spies. "No More Alice to Bob: Reality-based Models for Message Encryption and Key Management". Talk or presentation, 29, September, 2005.

Abstract
Communication security has long subsisted under a model motivated by the assumption that endpoints were secure, while intermediaries and third parties were untrusted. The natural implication of this model is that properties like non-repudiation, confidentiality and end-entity authentication be provided in an end-to-end fashion. Not only is this model and its implications incorrect in real systems, but it is actively detrimental to building systems that customers need. This talk will go into experiences integrating encryption into a major operating system, and also the realities of deploying email encryption within 100,000 user enterprises, and will attempt to distill a set of different security and design assumptions that lead to useful systems.

Electronic downloads

• spies_NoMoreAliceToBob.ppt · application/vnd.ms-powerpoint · 1148 kbytes

Citation formats

• HTML

  Terence Spies. <a
  href="http://www.truststc.org/pubs/7.html"
  ><i>No More Alice to Bob: Reality-based Models for
  Message Encryption and Key Management</i></a>,
  Talk or presentation,  29, September, 2005.

• Plain text

  Terence Spies. "No More Alice to Bob: Reality-based
  Models for Message Encryption and Key Management". Talk
  or presentation,  29, September, 2005.

• BibTeX

  @presentation{Spies05_NoMoreAliceToBobRealitybasedModelsForMessageEncryption,
      author = {Terence Spies},
      title = {No More Alice to Bob: Reality-based Models for
                Message Encryption and Key Management},
      day = {29},
      month = {September},
      year = {2005},
      abstract = {Communication security has long subsisted under a
                model motivated by the assumption that endpoints
                were secure, while intermediaries and third
                parties were untrusted. The natural implication of
                this model is that properties like
                non-repudiation, confidentiality and end-entity
                authentication be provided in an end-to-end
                fashion. Not only is this model and its
                implications incorrect in real systems, but it is
                actively detrimental to building systems that
                customers need. This talk will go into experiences
                integrating encryption into a major operating
                system, and also the realities of deploying email
                encryption within 100,000 user enterprises, and
                will attempt to distill a set of different
                security and design assumptions that lead to
                useful systems.},
      URL = {http://www.truststc.org/pubs/7.html}
  }
  

Posted by Christopher Brooks on 30 Sep 2005.
For additional information, see the Publications FAQ or contact webmaster at www truststc org..