Do I need end-to-end encryption on my mailbox?
Inbox Zero for Privacy, not Productivity inspired me to achieve an actual Inbox Zero at my e-mail provider, thereby sufficiently solving the challenge of e-mail encryption for my risk profile.1 As a result, I am able to choose a budget-friendly e-mail provider without paying a premium for 'privacy features'.
I chose to use the Thunderbird method as described in the blog post linked above. In practice, I rely on the following approach:
- Recently received e-mails: These e-mails remain in my Inbox and are accessible from any device (Webmail/IMAP) until I process them.
- Processed e-mails which I might require on the go (for example: reservations, hotel bookings, tickets): These e-mails get archived to my local Thunderbird archive when I am on my laptop. Moreover, I save a temporary copy (.eml) of these e-mails in a folder which I sync with Syncthing. As a result, I can access these e-mails on any device, even though they are no longer hosted at my e-mail provider.
- Processed e-mails which I do not / no longer require on the go (residual category): These e-mails are archived only to my local Thunderbird archive.
Whilst there are alternative approaches such as a self-hosted OfflineIMAP server, I find the above approach to be both reliable and low-maintenance. The only disadvantage is that I can only access my entire e-mail history from my laptop. That said, I consider the benefit of having (almost2) no exposure in the event of a data breach at my e-mail provider to vastly outweigh any such disadvantage.
If you are looking for more information regarding email encryption I can highly recommend the summary by Fastmail. Whilst you should always remain skeptical given that Fastmail has its own incentives, I do believe their post summarizes the current landscape rather well. I consider e-mail to be unreliable for actual end-to-end encrypted communication.↩
The approach described in this blog post does not decrease any ongoing exposure, such as an outstanding warrant or ongoing data breach at my e-mail provider. The approach only limits the scope of any breach to e-mails sent and received as of the date of the warrant or data breach, rather than my entire e-mail history (15+ years and counting).↩