Disk-level and file-level encryption is not supported at Cornell (including in ChemIT)

• No documentation.
• No recommendations
• No central key-escrow.
• No support for either technical staff nor end-users.

ChemIT's current capabilities

• Consultation regarding encryption
  • Risk management, whether encryption or its alternatives, is as much an issue of ones behaviors and attitudes as it is about technology.
  • Note: We personally have very limited experience with encryption.
• Benefits of encryption.
  • Use-cases when it's helpful. And when it's not helpful.
  • Encryption can represent unnecessary bother and risks in meeting some needs.
• Alternatives to encryption.
  • If the data is not on your laptop (for example), there is no need to have encrypted data on the laptop.
  • Identifying benefits and risks of various alternatives.
• Risks if using encryption. Currently, must self-escrow keys.
  
  • You may lose your data because of the encryption, through technical failures or losing keys.
  • Your backups may not be set up correctly relative to your encryption choices.
  • There is no university support for failed situations, including from ChemIT staff.
• Very limited training via some show-and-tell related to encryption or alternatives.
  • We cannot provide on-going support for anyone choosing to use encryption.

Future direction

If and when Cornell provides central support for file-level or disk-level encryption, ChemIT can naturally expand its services as appropriate.

• Coming as part of CU's MS Configuration Management service?

Continue to gauge demand within CCB. Requests for the service are almost non-existant so far. We recognize that although CCB use-cases seem low, but perhaps some risks are well hidden. (Faculty do travel. What data do they take with them?)

• Would those same, well-hidden use-cases avail themselves of an encryption service?

Possible, if deemed sufficiently valuable to CCB:

• Advocate for a central encryption service.
• Set up a service.
  • R&D, then testing promising solutions. Vetting can take a lot of time to uncover all the undocumented "gotchas". Develop recommendations, along with recommended alternatives.
  • Invest in, and bear the risks, associated with provisioning a local key escrow.
  • Invest in documentation and maintaining that documentation.
  • Provide in-person support for when things are going well. And for when things do not go well, when a higher technical competence is expected, often specialized.