Where work messages leak from
When a company investigates a leak, it most often turns out to be one of five things:
- Employees message each other in personal messengers, and work chats stay with the person even after they leave.
- A phone is lost or stolen, with a work chat open on it and no screen lock.
- The mail or messenger server stores messages unencrypted, and an administrator, a contractor or an attacker has access to it.
- A scammer writes in the name of a manager or a partner and asks for a document to be forwarded or an invoice paid urgently.
- An employee opens an attachment from a fake email, and malware gets onto the computer.
Each of these routes is closed in its own way, so a single measure such as a complex password is not enough.
Step 1. End-to-end encryption
The first rule: messages must be encrypted on the sender’s device and decrypted only on the recipient’s device. This is called end-to-end encryption. Then the server stores only ciphertext, and neither an administrator, nor a contractor, nor whoever hacks the server can read the messages.
Check that end-to-end encryption is always on, not just in “secret chats”, and that it covers groups, files, calls and mail. If even one of these features works without it, that is exactly where the leak will start.
Step 2. A work account, not a personal one
Work messages should live in a separate work account managed by the company. Then the administrator adds a new employee to the right groups in a minute, and when they leave, wipes the work account from their devices – the person’s personal chats remain untouched.
It’s important that the administrator manages access but has no keys to the content. Otherwise one compromised administrator account will open up all of the company’s messages.
Step 3. Device control
The phone is the weakest point. A minimum set of rules:
- screen lock and encryption of the device’s storage;
- the ability to revoke a device remotely if it is lost;
- a limit on the number of devices per person;
- a warning if the device shows signs of being compromised.
Think separately about messages that shouldn’t stay on the device: “View once” mode deletes them after viewing and, where the system allows, blocks screenshots.
Step 4. Protection against fake emails and files
Scammers count on haste. It helps when the app verifies the other person itself and shows that a message “from the boss” came from a new, unverified device. Unknown contacts should first be able to send only a request, without files. Attachments should be checked before opening, and programs and scripts in attachments blocked for everyone.
Step 5. Rules people will follow
A rule that gets in the way of work gets bypassed. So secure messaging has to be convenient: the same app on computer and phone, chats, files, calls and mail in one place, and it should work on low-cost phones. Then employees won’t want to go back to personal messengers.
Finally, the event log: who added a person to a group, who revoked a device, who changed the rules. The log shouldn’t contain message content, but it should answer the question “who and when”.
How CUBRA does it
CUBRA Messenger encrypts messages, files, calls and mail end to end with two independent layers and post-quantum keys. Work accounts, policies and device revocation are in CUBRA Console, which has no access to content. The communication server can be placed on your own premises – this is CUBRA Node.