Thoughts

- Reading time: ca. 8 minutes -

Thoughts on “end-to-end encryption”

This topic fills library shelves - however, these thoughts are not intended to provide instructions for action or suggested solutions, but rather to be a stimulus for further thought.

Before talking about end-to-end encryption, however, “security” must be briefly discussed, as the two are related. However, security/pseudo-security is such a large and important topic that there are separate thoughts on this - here is the brief summary:

End-to-end encryption

“End-to-end encryption” is a method of transmitting encrypted data in networks. In this case, the message encrypted by the sender is transmitted unchanged even over several computers to the recipient. It is often abbreviated to “E2E” or “E2EE”.

Intended use

On the one hand, the use seems complicated - yes. On the other hand, however, end-to-end encryption is very good and useful, because examples/arguments exist for both business and private purposes:

  • “I want E2EE because I don’t run the server myself.”
    (If you don’t want to let strangers access the content)

  • “I want E2EE because I run the server myself.”
    (In a family where one family member is also the server administrator and without encryption could read the messages of the other family members and doesn’t want that)

So there is also a lot to be said for this (except convenience). In principle, everyone should use end-to-end encryption if it makes sense - but only if the principle is understood.

Verification

Regardless of whether encryption is enabled or not, the biggest vulnerability and the weakest link is usually the end device in question or the person operating it. And end-to-end encryption only makes sense if the respective keys are located exclusively on the end devices. If the private keys are on servers, it is not even necessary to build backdoors into programs to spy out data unnoticed.

Unfortunately, many do not understand that with true end-to-end encryption, each device used must be verified by the others - this is important!

Exactly this required verification is often concealed by manufacturers/vendors and often not done by users. The main thing in the brochure or in the chat is “end-to-end encrypted”: Then everything is good - and everything else is suddenly no longer important and is suppressed.

In addition, “E2E” has increasingly become a main argument and marketing tool in recent years (including by WhatsApp/Facebook). One could get the impression that the hype around “E2E” encryption was caused by the messenger industry and that they are trying to suggest to people “You absolutely need this to live!”.

But those who loudly tout “secure” encryption can and often do blind others, thereby barely noticing and incidentally gaining valuable metadata. In other words, this pushes data protection and privacy into the background.

On the other hand, the manual non-verification of keys then corresponds to the “TOFU” principle (trust on first use). And this has a charm all of its own: If no attacker interfered during the initial connection setup, then everything works very well without manual effort. This is definitely sufficient for normal security requirements - but who knows what normal is.

Does a WhatsApp user really expect and need real and tested end-to-end encryption? Do they really need it or do they want it?

Transport encryption

Independently of “end-to-end encryption”, there is also “transport/line encryption” (or “point-to-point encryption”). With this transmission method, the message is only encrypted for the respective neighboring computer. This method is also used, for example, when calling up “secure” Internet pages.

Both transmission methods have strengths and weaknesses - but they can also be used in combination. A pictorial description of the differences as well as the combination options follows below.

Is end-to-end encryption overrated?

Why are business documents, confidential and sensitive documents, etc. often sent by e-mail without encryption (e.g. PGP/GPG (external)), and why is “forward-looking security” and “deniability” suddenly so important for chatter (chat)? Wasn’t there something? That’s right:

Everyone “have to have” end-to-end encryption!

The main thing is that it says “encrypted” on it. Convenience and pseudo-security are more important to many - or it is simply due to ignorance.

In this context, however, one must not forget that there are other criteria for the choice of a messenger (system) besides “encryption security”, which can be evaluated/weighted individually and differently:

  • Freedom,
  • privacy,
  • independence,
  • privacy,
  • security,
  • range of functions,
  • costs, …

With this knowledge, the question …

“Is end-to-end encryption overrated?”

… can be answered: Yes, often!


Encryption types and possible combinations

„Explanation

The explanation graphic can also be downloaded as a print file: Verschlüsselung.PDF (approx. 0.4 MB)

Attack scenarios

As far as “security” is concerned, true (verified) end-to-end encryption can only be a protection against very specific attack scenarios.

And unverified end-to-end encryption, which is most commonly used in practice, covers even fewer scenarios - namely, only this one:

  • “attackers” are not (only) interested in metadata
  • but (also) for content,
  • has read access to the content but
  • no write access to the content and
  • no access to the end devices.

And that in this respect the typical marketing of closed messenger systems (“data silos”) such as WhatsApp, Signal, Threema & Co. can be misleading if it is suggested that (let’s call it “fake”) end-to-end encryption is the decisive aspect for evaluating solutions.

  • Without manual checking (verification), E2EE only protects against passive attacks - third parties want to read along unnoticed.
  • With manual checking (verification), however, it also protects against active attacks - a third party wants to get in/between the communication without the parties involved noticing; “MITM” attack (“man in the middle”).

No end-to-end encryption

There are also cases where end-to-end encryption (E2EEE) does not make sense and transport encryption is completely sufficient.

In the business environment, for example, the keywords “substitution”, “emergency”, “employee change”, “backup” or even “traceability” are indications of such use cases.

For example, what does a company do when important business correspondence can only be opened by “terminated” or “former” employees? There have been cases where an important password was simply forgotten or accidentally changed before the employee left the company.

In the private environment, many savvy users do not practice “real” (verified/authenticated) end-to-end encryption for the majority of their contacts in the meantime, and only do actual key matching for a few contacts:

Those who have understood that manual verification of each device/client is required and, for example, a device change naturally has corresponding consequences. After all, from which “attack” does one want to protect oneself and with which effort?

Summary/Conclusion

End-to-end encryption is a good tool and often useful - but: Die Ende-zu-Ende-Verschlüsselung ist ein gutes Werkzeug und oft sinnvoll - aber:

The question “Is end-to-end encryption overrated?” can be answered with a “YES!”.

As a partial aspect of security, end-to-end encryption is important, but often misunderstood or overrated!

Supplementary information:

Date: 22.04.2022
Rights: CC BY-SA
Autors: Diverse (Initiative Freie Messenger)


All articles/thoughts about Messenger: