Who Owns Our Conversation? – Part 2

February 6, 2026

Exploring alternatives to big‑tech and why non‑business structures matter

Quick recap of Part 1

If you missed the first installment, you can read it here:
Who Owns Our Conversations: Testing the Fediverse

In Part 1 we examined the Fediverse as a decentralized counter‑weight to the mainstream social web. In this sequel we turn our attention to services that start as small projects or startups and later get absorbed by large corporations. Those acquisitions often rewrite the service’s original promise, turning a privacy‑first offering into a profit‑driven machine. The pattern leaves users scrambling for trustworthy alternatives that stay true to the values that attracted them in the first place.

When “small” becomes “big”

Many of the platforms we rely on daily began as modest ventures hoping to give users a better experience. After gaining traction, they attracted the attention of big tech, which bought them out or forced a strategic partnership. The result is usually threefold:

Values shift – promises like “no ads” or “no data selling” disappear or become heavily qualified.
Monetisation pressure – the new owners look for ways to monetise the user base (ads, data licensing, AI training, or cross‑selling other products).
Feature creep – new features are introduced that serve the parent company’s ecosystem rather than the user’s needs.

These dynamics create a vacuum: users who care about privacy, transparency, or independence start looking for services that are structurally insulated from such pressures.

The structural difference: non‑business & hybrid models vs business

A non‑profit foundation or a foundation‑backed hybrid company can act as a firewall against hostile takeovers and profit‑centric pivots. The key idea is that the entity that owns the service does not answer to shareholders demanding ever‑higher growth rates. Instead, its mission is codified in a charter or bylaws that prioritise user rights, privacy, and long‑term sustainability.

Because the governing body’s incentives differ, the technical and policy choices it makes tend to stay aligned with the original promise:

Funding comes from donations, grants, or a modest paid tier rather than advertising revenue.
Roadmaps are publicly discussed and often include community input, reducing the chance of sudden, user‑unfriendly changes.
Legal structures (foundations, cooperatives, or non‑profits) make it considerably harder for a venture‑capital‑backed acquirer to simply buy the service and flip its policies overnight.

Below are two separate tables that illustrate how this structural difference plays out in practice for messaging and email services.

Messenger comparison

Feature	Signal	WhatsApp	iMessage	Telegram
Ownership model	Non‑profit foundation (Signal Foundation)	Subsidiary of Meta Platforms, Inc.	Proprietary service owned by Apple Inc.	Private company (Telegram Messenger LLP)
Core privacy claim	End‑to‑end encryption + zero‑knowledge architecture – the service never sees plaintext	End‑to‑end encryption for one‑to‑one chats, but metadata is stored on Meta’s servers; group chats are not E2EE by default	End‑to‑end encryption only for iMessage (Apple‑to‑Apple); SMS fallback is unencrypted	Server‑client encryption (MTProto); optional “Secret Chats” with E2EE, but most chats are stored in the cloud
How the model protects users	No shareholders; funding via donations keeps the product free and focused on privacy	Meta’s ad‑driven business model incentivises data collection and integration with its broader advertising ecosystem	Apple’s ecosystem limits data sharing, but the service is tied to a single hardware vendor and can be subject to court orders in jurisdictions where Apple complies	Telegram’s revenue comes from optional premium features; the company has historically resisted providing data to governments, but the default cloud storage means the provider could access content if compelled

Key advantage of Signal: because it is run by a non‑profit foundation, there is no pressure to introduce advertising, data‑licensing, or feature bloat that would compromise its privacy‑first stance. Its funding model (donations and grants) aligns directly with the mission of keeping communications private for everyone.

Email comparison

Feature	Proton Mail	Gmail	Outlook.com	Yahoo Mail
Ownership model	Hybrid: Proton Foundation (non‑profit) + commercial arm for paid plans	Subsidiary of Alphabet Inc. (public corporation)	Subsidiary of Microsoft Corp. (public corporation)	Subsidiary of Verizon Media (public corporation)
Core privacy claim	End‑to‑end encryption + zero‑knowledge storage – Proton cannot read your emails	Transport‑level TLS encryption; emails are stored in cleartext on Google’s servers and can be processed for AI features (e.g., Smart Compose)	TLS encryption in transit; emails stored in cleartext on Microsoft’s servers; data used for personalised services	TLS encryption in transit; emails stored in cleartext; data used for advertising and personalization
How the model protects users	The foundation’s charter forbids data mining; paid plans fund the free tier while preserving privacy	Google’s business model relies on data aggregation; while ads are no longer based on email content, the service still powers AI models that learn from the corpus	Microsoft integrates email data with its broader Graph ecosystem, enabling targeted services and ads for free accounts	Yahoo leverages email data for ad targeting and sells aggregated insights to third parties

Key advantage of Proton Mail: the Proton Foundation’s mission‑driven charter prevents the service from mining email content for advertising or AI training. Even though a commercial arm offers paid plans, the revenue stays within the ecosystem and does not alter the core privacy guarantees.

Why the technical details matter, even if you don’t notice them

End‑to‑end encryption (E2EE) – Messages are encrypted on the sender’s device and can only be decrypted on the recipient’s device. The provider never holds the decryption keys, which means a court order or a data breach cannot expose the plaintext.
Zero‑knowledge architecture – The service stores only ciphertext. Because it never possesses the keys, it cannot read, sell, or repurpose your data, even if compelled by law.
No data‑selling pipelines – Without access to plaintext, a provider cannot build detailed user profiles for advertising or sell data to third parties.

If you want a deeper dive into how encryption works and why it matters, the Electronic Frontier Foundation offers an excellent primer: https://ssd.eff.org/module/what-should-i-know-about-encryption#end-to-end-encryption.

Advertising, data mining, and government requests

Advertising & data mining – Gmail and Outlook do not scan email content to serve personalised ads (Google stopped that practice for free accounts in 2017). However, both services analyse metadata and content to power AI‑driven features such as Smart Compose, spam filtering, and predictive search. Those analyses can be leveraged for product improvements and, indirectly, for advertising ecosystems.
Government requests – In many jurisdictions, authorities can subpoena email metadata or, in some cases, request content. With end‑to‑end encryption, even a valid court order yields only metadata, not the message bodies. Services that lack E2EE (e.g., standard Gmail or Outlook) can be forced to hand over the actual content, sometimes years after the communication took place. This risk is especially acute for journalists, activists, and anyone living under authoritarian regimes where the definition of “national security” can shift dramatically.

Having a service that implements both E2EE and zero‑knowledge storage therefore provides a future‑proof safeguard: even if the legal environment changes, the provider simply does not have the ability to comply with a request for the plaintext.

Owning the conversation

Owning the conversation is not about finding the perfect app or the purest platform. It is about understanding the structures behind the tools we use and being honest about the trade-offs we accept when we use them. Convenience is never neutral, and neither is scale.

Decentralization, foundations, and open protocols are not guarantees of freedom, but they do something essential: they slow down the moment where incentives quietly change. They introduce friction where power would otherwise concentrate unnoticed. That friction matters.

The question, then, is not whether a service encrypts messages or publishes a transparency report. The real question is who can change the rules tomorrow, and under what pressure. Growth, investors, geopolitics, and regulation all shape those answers long after the onboarding screen fades away.

We often talk about speech as if it exists independently of infrastructure. It doesn’t. Conversations live inside systems, and systems always have owners — legal, economic, or political. Choosing where we speak is, whether we like it or not, choosing who ultimately has leverage over that space.

Knowing who owns the conversation is the first step in deciding how much of ourselves we’re willing to share and how much we want to keep, and how we’ll protect it.

Billionaire acquisition of TikTok – An example of a high‑profile deal that illustrates how a popular platform can change hands and shift its policies. https://www.floridatoday.com/story/news/2026/01/28/who-bought-tiktok-florida-billionaire-oracle/
What you should know about encryption – The EFF’s guide to end‑to‑end encryption and zero‑knowledge concepts. https://ssd.eff.org/module/what-should-i-know-about-encryption#end-to-end-encryption
Signal – official site – The non‑profit messenger that exemplifies the privacy‑first model. https://signal.org/
Proton Mail (affiliate link) – Supports the free tier while keeping the service independent. https://go.getproton.me/SH2L3