Encrypted in transit is not the same as end-to-end
When you send a message, it passes through several systems on its way: your device, your network, the messaging company’s servers, and finally the recipient’s device. Protection can be applied in different places, and the differences matter.
With ordinary transport encryption, the connection between your device and the company’s server is scrambled, so someone watching your Wi-Fi cannot read it. But the company’s servers can read the message, because they decrypt it, store it and pass it on. The protection is against outsiders on the network, not against the service itself.
With end-to-end encryption, the message is scrambled on your device and can only be unscrambled on the recipient’s device. The company’s servers carry the data but cannot read it. The “ends” are the two people (or devices) in the conversation.
How end-to-end encryption works in outline
Most systems use public-key cryptography. Each person’s device generates a pair of mathematically linked keys: a public key that can be shared freely and a private key that never leaves the device. A message locked with someone’s public key can only be unlocked with their matching private key.
In a typical secure messaging app:
- Your device obtains your contact’s public key, usually through the company’s server.
- Your app uses it, along with your own keys, to agree on a shared secret that only the two devices know.
- The message is encrypted with keys derived from that secret and sent.
- The recipient’s device uses its private key material to decrypt it.
Well-designed apps also change the encryption keys regularly as a conversation continues, so that discovering one key does not expose the whole history or all future messages. Group chats use related techniques with extra care for members who join and leave.
The details differ from app to app, and the strength of a particular product depends on how carefully it was designed and implemented. Independent review of the design helps, but it cannot make guarantees.
Checking you are talking to the right person
There is one weak point in any key-exchange process: you need to be sure the public key you received really belongs to your contact. If an attacker could swap in their own key, they might be able to sit in the middle. Many apps therefore offer a way to verify, such as comparing a short code or scanning a QR code in person or over a separate channel. Verifying is optional but worthwhile for particularly sensitive conversations. Many apps also notify you when a contact’s key changes, which can have innocent causes like a new phone but is worth noticing.
What stays visible: metadata
Even when the content is hidden, other information about the conversation often remains available to the service, and sometimes to others. This is called metadata, and it can include:
- who is communicating with whom, and how often
- when messages are sent and how large they are
- your phone number, email address or username, and profile details
- your device type, IP address and approximate location
- the groups you belong to, depending on how the app is designed
Different apps collect and keep different amounts of metadata, and some deliberately minimise it. Metadata can say a lot about someone’s life even without any message text, which is why it is worth reading an app’s privacy policy and understanding what it collects. The app’s own description of its practices is the best starting point, and claims are worth treating with some caution until you understand them.
How backups can weaken the protection
This is one of the most common gaps. End-to-end encryption protects a message while it is travelling. But the copy stored on your phone can be saved elsewhere, and that copy may not be protected in the same way.
If your messages are backed up to a cloud account in a form the provider can read, then the provider, or anyone who gains access to that account, may be able to read your history, even though the messages were encrypted in transit. A chat is also only as private as the other person’s phone and backup settings: you can protect your side perfectly and still have your messages stored in an unprotected backup by the person you wrote to.
Some apps now offer encrypted backups protected by a password or key that only you hold. That removes the gap but adds a responsibility, because if you forget the password the backup may be unrecoverable. Check your app’s settings to see what it does by default and which options exist.
Other limits to keep in mind
- Screenshots and forwarding. The recipient can copy, screenshot or share what you send. Disappearing messages are a convenience, not a guarantee, since nothing stops a person or another device from saving the content.
- Compromised devices. If someone controls your unlocked phone, or malware is running on it, the messages are readable at the endpoint. Keeping software updated and using a screen lock are part of the picture, and so is device encryption for the stored history.
- Account takeover. If an attacker takes over your account, for example through SIM swapping, they may be able to impersonate you going forward. Turning on any available registration lock or PIN adds some protection.
- Notifications. Message previews on a lock screen can reveal content to anyone who sees the phone.
- Impersonation. Phishing scams arrive through secure apps too. Encryption does not tell you whether the sender is who they claim to be.
Choosing sensibly
When comparing messaging apps, useful questions are whether end-to-end encryption is on by default for all chats or only some, what metadata is kept, whether backups are protected, and whether the design has been publicly reviewed. A secure messaging app protects the conversation in transit very well when it is built properly, but protecting the conversation as a whole depends on the devices, backups and habits around it.