There is an image that repeats.

Someone sits in a café. Four people at the next table, all mid-twenties. Nobody speaks. Everybody types. Four displays, four apps, four streams of messages flowing into four different data centres — past servers none of them has ever seen, companies none can name, under conditions none has ever read.

None finds this strange. That is the state we have reached. And it is not normal — it is merely familiar.

A brief history of messaging

Human communication is old. Writing is old. Letters are old. But electronic messages — the idea of turning thoughts into bits and sending them over wires — is young. And its history is a tale of tools that came and went while one constant remained.

Email. In 1971 Ray Tomlinson sent the first network message between two computers sitting beside each other. He chose the @ symbol because it sat on the keyboard and had no other use. What began as a technical experiment became the longest-lived communication instrument of the digital age. Email is not sexy. It is not elegant. It has problems — spam, abuse, phishing, absent authentication in the original design. But it endured. Why? Because it rests on open standards. SMTP, IMAP, POP3 — protocols owned by nobody. Anyone can run a server. Anyone can speak to anyone. No vendor can shut email down or pull it from circulation. Email is federated by nature — and that is exactly why it still exists while dozens of successors came and vanished.

IRC. 1988. Real-time chat for academic networks. Channels, nicknames, linked servers. Open, federated, simple. It still exists — not because it is modern but because it works and belongs to nobody.

ICQ. 1996. The pager for the internet. ICQ was the first mass messenger. Everyone received a number — the famous UIN. You collected contacts, sent short messages, saw who was online. ICQ was a phenomenon. It defined a generation. And then it vanished. Not overnight, but steadily. First came AOL, the acquisition, integration. Then came competitors that were better, more modern, mobile-first. ICQ was shut down in 2024, after nearly three decades. A tool that shaped millions no longer exists. Where did the conversations go? They disappeared. Because ICQ was a central service. One provider. One server. One owner. When the owner decided it was over, it was over.

XMPP/Jabber. An open standard for instant messaging, federated like email. Google Talk used XMPP — then turned federation off. Facebook Chat used XMPP — and did the same. Both companies leveraged the open protocol to gain users, then slammed the doors once network effects were large enough. That is no coincidence. That is strategy.

De-Mail — the failed attempt

Germany launched De-Mail — a state-certified system for binding electronic communication. The idea: email, but legally binding, with identity verification, transport encryption, delivery proofs.

It failed. Not on technology. On the business model. De-Mail was complicated. It cost money. It was bound to providers. It offered nothing people wanted — it offered what authorities wanted. And it arrived late, in a world already accustomed to messengers that were faster, simpler, mobile-first. De-Mail formally still exists, but nobody uses it. It is a monument to the failure of centralised, state-run communication solutions that forgot the user.

The present: popularity as business

Now we talk about the essentials.

WhatsApp. Telegram. Threema. Signal. Matrix. Five messengers, five models, five answers to the question: whom do you trust with your conversations?

WhatsApp has over two billion users. It is the de facto standard. When someone says “I’ll message you” they mean WhatsApp. And WhatsApp belongs to Meta — the company that also owns Facebook and Instagram. The same company that through Cambridge Analytica proved it does not merely tolerate data abuse but operates it as a business model.

WhatsApp offers end-to-end encryption. That is good. But encryption protects only content. Metadata — who speaks with whom, when, how often, from where — is captured. And metadata reveals more than content. Whoever knows the contact networks knows the social structure. Whoever knows the timing knows the rhythm. Whoever knows locations knows movements. Content is one line. Metadata is the whole book.

Telegram has over nine hundred million users. Telegram advertises freedom and security. But Telegram does not encrypt group chats end-to-end by default. Secret Chats must be activated manually — and work only between two devices. Standard chats run on Telegram’s servers, readable in plaintext by the company. Telegram has handed metadata to authorities in the past, proving it cooperates under pressure. Whoever considers Telegram a safe choice has not looked closely.

Systematic surveillance

Here it gets uncomfortable. Because this is no longer about individual firms collecting data. It is about structures that surveil people systematically.

Popular messengers are not mere communication tools. They are sensors. Every message, every contact, every location, every file is a data point. And these data points converge — not in a single corporate silo but in an ecosystem of authorities, intelligence services, data brokers and AI systems that build profiles from them.

ChatControl — officially “Regulation to combat child sexual abuse” — is the most prominent example. The idea sounds honourable: protect children. The means are menacing: all messages — emails, chats, files — are to be scanned automatically. Including encrypted ones. To do so, encryption must be broken or bypassed. Client-side scanning is the magic word: the scanning software runs on the user’s device before the message is encrypted. What is supposed to be private is searched on your own device — by software you do not control, according to criteria you do not know, with error rates nobody guarantees.

The problem is not the intention. The problem is the architecture. Whoever has installed a backdoor once has installed it for everyone. For authorities wishing to use it legitimately. For intelligence services that do so anyway. For hackers who find it. For foreign states demanding it. A backdoor is not a lock with two keys. It is a hole in the wall — and holes cannot be controlled.

Whoever consents to ChatControl because they wish to protect children effectively consents to the end of private communication. Because the scanning system built once for illegal imagery can be reconfigured for arbitrary content. Political opinions. Religious affiliations. Union memberships. Protest plans. Anything a state someday defines as suspect becomes scannable — because the infrastructure already exists.

Why people are careless

Why do two billion people use a messenger owned by a data conglomerate? Why do nine hundred million use one that does not encrypt what it pretends to protect?

Because it is convenient. Because everybody else does. Because switching takes effort. Because consequences are invisible.

Data abuse does not hurt. Not immediately. When Meta sells your profile to advertisers you do not feel it. When an algorithm analyses your traffic to predict purchasing behaviour you do not notice. When a state collects metadata to identify dissidents you learn it — if at all — only when it is too late.

Carelessness is not stupidity. It is adaptation to an environment where the cost of vigilance seems higher than the cost of disclosure. But that is an illusion. The cost of disclosure pays off not immediately. It pays off when you become politically active. When you blow the whistle. When you become a target. When circumstances change and suddenly the data you casually left behind is used against you.

Whoever does not choose, chooses nonetheless. They choose the status quo. And the status quo is: maximal data capture, minimal control.

Why deeper knowledge is needed

A good choice requires knowledge. Not superficial. Deep.

You must understand what end-to-end encryption means — and what it does not. It protects content. It does not protect metadata. It does not protect against malware on the device. It does not protect against keylogging. It does not protect when the vendor controls key distribution and silently redirects it.

You must understand what Perfect Forward Secrecy means — that recording all traffic today does not help if the key is compromised tomorrow. You must understand what federation means — that you can speak to anyone regardless of server. You must understand what self-hosting means — that you control the servers processing your conversations.

And you must understand how cryptography can be exploited. Client-side scanning is an attack on encryption from within. Key escrow — depositing keys with third parties — is a backdoor sold as a service. “Lawful interception” is the euphemistic term for systematically breaking encryption on demand.

Without this knowledge every choice is blind. And blind choice in a world where data is power is not a choice — it is surrender.

Why it is time to delete WhatsApp

Here it gets concrete. And more personal.

Deleting WhatsApp is possible. It takes resolve, but it is doable. You lose contacts — but you regain something. You signal: my conversations belong to me. My data is not self-evident. My choice is conscious.

Signal is the minimally invasive alternative. Open source, funded by donations, without a data business model, with robust end-to-end encryption, encrypted profile data, minimal metadata capture. Signal is not perfect — it is centralised, bound to one server fleet, requires a phone number. But it is the minimum anyone serious can do. Using Signal instead of WhatsApp marks the most important decision: the decision to decide at all.

Matrix goes further. Matrix is federated like email. Anyone can run a server. Anyone can speak to anyone — regardless of homeserver. Matrix encrypts end-to-end by default. Matrix is open source. Matrix is self-hostable. Matrix is the model that email proved four decades ago — federated, open, owned by nobody — applied to real-time communication.

Matrix is superior. Not because it has the largest feature set. But because it is the architecture that enables sovereignty. Whoever runs their own Matrix server controls where conversations are processed, who sees metadata, who has access. Whoever relies on a foreign server trusts. Whoever self-hosts knows.

Consistency counts

A choice is only a choice if you stick with it. Whoever uses Signal but keeps WhatsApp “just for those contacts” has decided nothing — they maintain two silos instead of one. Whoever advocates Matrix but replies on Telegram undermines their own position.

Consistency is what turns a preference into a conviction. It means being inconvenient. Saying: “I’m no longer on WhatsApp. Message me on Signal — or by email.” It means some conversations will drop away. Some people will lose touch. Some comforts vanish.

But it also means beginning to shape the world the way you want it — not the way it is presented to you. Everyone who switches consistently makes switching easier for others. Everyone who stays with their choice makes the choice visible. And visibility is the beginning of momentum.

How libcom.de can help

Here it gets practical. Especially in corporate settings.

Choosing a messenger is a personal decision in private life. In a company it is strategic. Whoever handles employee communications over WhatsApp groups delivers internal information to Meta. Whoever conducts customer conversations over Telegram stores trade secrets on Russian servers. Whoever coordinates projects over Microsoft Teams ties communication to a vendor that dictates licence prices.

libcom.de advises organisations on choosing and introducing sovereign messaging solutions. We have experience deploying Matrix — from small team servers to federated enterprise infrastructures. We plan, implement, document and train. We help migrate from WhatsApp, Telegram, Teams to Signal or Matrix — without operational disruption, without data loss, with training concepts that enable adoption.

We do not start with a product but with an analysis: how do you communicate today? What are your requirements — legal, technical, organisational? What dependencies exist? What risks are acceptable, which are not? Then we plan — in stages, documented, reversible. And we are honest when a solution does not fit your case.

If you wonder whether your corporate communication is sovereign — or whether it was long ago delegated to corporations: write to us at contact@libcom.de. We make an honest inventory. Without sales pressure. With a view to what lasts long-term.

The history of messengers teaches: what is central disappears. What is open endures. ICQ is gone. Email exists. XMPP was used by Google and Facebook and then betrayed. De-Mail was ignored. WhatsApp belongs to Meta. Telegram does not encrypt what it pretends.

The choice is not hard because the options are unclear. The choice is hard because the consequences are inconvenient. But it is the most important choice we make daily — the choice of whom we entrust with our voice.

Signal is the minimum. Matrix is the goal. And the first step is to choose at all.