How Europe Is Quietly Ditching US Software, Country by Country

Something significant is happening across Europe, and it is not making front-page headlines in most places. Quietly, methodically, and with increasing urgency, European governments and companies are replacing American software with European alternatives. Not just one country, and not just one product. The shift is happening across operating systems, office suites, video conferencing tools, messaging apps, and now AI assistants.

This is not technophobia. It is strategy.

The past two years have brought a series of wake-up calls: cloud outages affecting millions of users, US companies routing EU data abroad without asking, AI providers signing compute deals with defense-linked American entities, and a US administration openly hostile to European interests. Combined with the CLOUD Act, which allows American authorities to demand access to data stored on any US-company server anywhere in the world, European governments have concluded that digital dependency on American infrastructure is no longer acceptable.

Here is what they are replacing, what they are switching to, and why it matters to you.

Germany: LibreOffice, Linux, and Saying Goodbye to US Productivity Software

No country has moved further or faster than Germany in replacing US productivity tools with open-source alternatives. The story begins in Schleswig-Holstein, Germany’s northernmost state, which began one of the most ambitious public-sector migrations in European history.

Schleswig-Holstein’s migration to LibreOffice has reached 80% completion, with a one-time €9 million investment earmarked for 2026 to finish the job. Starting in 2026, the state will save over €15 million annually in license costs it previously paid to American software vendors. Beyond the finances, the reasons are structural. By pivoting to open source and hosting services locally in Europe, Schleswig-Holstein is effectively eliminating risks around data leaving the EU. Running Nextcloud on German soil removes the risk of data flowing outside EU jurisdiction.

The full stack being deployed is remarkable in its scope: LibreOffice for documents, Thunderbird for email, Open-Xchange for calendar and groupware, Nextcloud for cloud storage, and KDE Plasma on Linux as the desktop environment. Schleswig-Holstein successfully migrated its state administration email system, comprising more than 44,000 mailboxes and 110 million emails and calendar entries, to Open-Xchange and Thunderbird.

At the federal level, the direction is equally clear. In the spring of 2026, the German Federal Ministry for Digital Transformation and Government Modernization released a framework stating that in the future only open formats such as Open Document Format (ODF) will be permitted for official documents. This effectively ends the dominance of proprietary US document formats in German public administration.

What they switched to: LibreOffice (open source, developed partly in Germany), Linux (KDE Plasma), Nextcloud, Thunderbird, Open-Xchange.

France: Visio Instead of Teams and Zoom, Linux on Civil Servant Desktops, Olvid Instead of WhatsApp

France is executing the most comprehensive digital sovereignty strategy of any European country, touching video conferencing, operating systems, and messaging simultaneously.

Video Conferencing: Out with US Platforms, In with Visio

France will replace US video conferencing platforms, including Microsoft Teams and Zoom, with its own domestically developed tool, Visio, across all government departments by 2027. The stated goal is to “end the use of non-European solutions and guarantee the security and confidentiality of public electronic communications.” France’s civil service minister explained that the government had become “excessively dependent” on overseas tools and that ending their use was essential to protect internal discussions.

This is not a pilot project. Visio is already live with 40,000 active users, and major institutions including the CNRS are scheduled to transition 34,000 employees by March 2026. The Ministry of Defense and the National Health Insurance Fund are among the early adopters, signaling that this covers critical national infrastructure, not just peripheral agencies.

Visio is hosted by Outscale, a subsidiary of Dassault Systemes, which holds the SecNumCloud qualification, France’s highest security certification for cloud providers. This ensures that government conversations remain under strict French legal protections, immune to extraterritorial subpoenas from US authorities. The AI transcription features built into Visio use technology from French startups Pyannote and Kyutai, keeping even the AI layer within European jurisdiction.

The government estimates switching to Visio could save up to €1 million per year for every 100,000 users who move off commercial licenses. “We cannot risk having our scientific exchanges, our sensitive data and our strategic innovations exposed to non-European actors,” France’s civil service minister stated publicly.

What they switched to: Visio (built on open-source Jitsi, hosted on French sovereign cloud infrastructure certified SecNumCloud).

Operating System: France Plans Full Linux Migration

In April 2026, France announced its plan to move all government computers, those of 2.5 million civil servants, from proprietary US operating systems to Linux. This would be the largest government Linux migration in history if completed. The announcement follows years of partial moves, including the French Gendarmerie’s long-running Linux deployment covering tens of thousands of desktops.

Messaging: WhatsApp, Signal, and Telegram Banned, Olvid Mandatory

France was ahead of most countries in recognizing that consumer messaging apps, wherever they are headquartered, are a sovereignty problem. The French Prime Minister issued a directive banning the use of WhatsApp, Telegram, and Signal by ministers and their staff, citing security risks. The directive noted that these apps have “known vulnerabilities” and “do not guarantee the confidentiality, integrity and availability of the data exchanged.”

Government officials were instructed to switch to Olvid, a secure messaging app certified by the French National Cybersecurity Agency (ANSSI). Olvid offers end-to-end encryption without requiring access to users’ phone numbers or contact lists.

Why Olvid specifically? Unlike Signal, which is American and subject to US legal reach, Olvid is French. Critically, it encrypts not just message content but also metadata, meaning that even Olvid itself cannot determine who is communicating with whom. Cabinet members can also use Tchap, a messaging app developed by the French government itself, launched in 2019 as a secure communication platform for public servants.

What they switched to: Olvid (French, ANSSI-certified, encrypts metadata), Tchap (French government-built).

Denmark: Breaking the Big Tech Habit for Sovereignty Reasons

In summer 2025, Denmark’s Ministry of Digital Affairs announced that all its employees would work with Linux and LibreOffice instead of US software. The minister explained: “We must never make ourselves so dependent on so few that we can no longer act freely. Too much public digital infrastructure is currently tied up with very few foreign suppliers. This makes us vulnerable.”

Cities including Copenhagen and Aarhus followed suit. The geopolitical context matters here: US President Trump’s repeated claims regarding Greenland, Danish territory, made Danish officials acutely aware of what it means to depend on American technology companies when American political intentions toward your country are openly hostile.

What they switched to: LibreOffice, Linux.

Austria: Open Source in the Military, a National Sovereignty Charter, and European Messaging

Austria’s military has already switched from US office software to LibreOffice across its armed forces. Austria also initiated a national charter on digital sovereignty, which was adopted by all 27 EU member states during the Summit on European Digital Sovereignty in late 2025, with the goal of strengthening Europe’s autonomy over digital infrastructure by promoting open-source alternatives.

Switzerland, though not an EU member, has taken an equally firm line on messaging: the Swiss military ordered soldiers to stop using WhatsApp, Signal, and Telegram and to switch instead to Threema, a Swiss-developed encrypted messaging service. The reasoning is simple: Swiss law cannot protect Swiss soldiers if their messages flow through servers in US jurisdiction.

What they switched to: LibreOffice, Threema

Italy: The Ministry of Defence Leads on Open Source

Italy’s Ministry of Defence replaced US productivity software with LibreOffice on 5,000 machines as part of a broader effort to transition to open platforms. Italy’s move is particularly significant because defense ministries are among the most sensitive users of productivity software, handling classified procurement, strategy documents, and inter-agency communications. If open source is good enough for that environment, the argument for keeping proprietary US software in civilian agencies weakens considerably.

What they switched to: LibreOffice.

The Netherlands and the International Criminal Court: A Signal for Global Institutions

The International Criminal Court announced in October 2025 that it would switch from US office software to openDesk, a German open-source productivity suite, following US sanctions that led to the ICC’s chief prosecutor being disconnected from his American-hosted email account. This episode made painfully visible what European institutions operating in the international legal sphere cannot afford: communications infrastructure controlled by a company subject to US government pressure.

Amsterdam published a Multiyear Digital Autonomy Strategy covering 2026 to 2035, outlining a full path to digital sovereignty. Several Dutch ministries have also begun reconsidering their reliance on US collaboration tools amid strategic dependency concerns.

The Messaging Question: Why Popular Apps Are Not Enough for Europe

It is worth pausing on messaging, because many Europeans assume that Signal, being end-to-end encrypted and non-commercial, is a safe choice. Multiple European governments disagree, and their reasoning is sound.

Signal is an American company, incorporated in the United States, subject to US law. While Signal’s architecture means it holds very little user data, it is still subject to legal demands from US authorities, and its infrastructure runs on US servers. For government communications, “very little data” is not the same as “no data.”

European governments are looking to move away from “platforms over which they have no control.” This logic applies equally to WhatsApp, owned by Meta and deeply integrated into American surveillance capitalism, and to Telegram, which despite its reputation is incorporated outside Europe and has faced serious transparency questions.

The alternatives being adopted share a set of properties that matter specifically to European institutions: European incorporation, European hosting, subject only to European law, and in most cases audited by European cybersecurity agencies.

Wire, incorporated in Switzerland, offers end-to-end encrypted messaging and calling built on open-source protocols. It implements the Messaging Layer Security standard and is designed for enterprise-scale deployment with full administrative control. Swiss incorporation means it is not subject to US CLOUD Act demands. Wire is used by organizations across Europe that need secure team communication without routing data through American infrastructure.

Olvid, certified by France’s ANSSI, goes further than almost any other messaging app by encrypting metadata as well as content. A user’s contact list, conversation partners, and message timing are all protected. Olvid does not require a phone number to register, which is significant: WhatsApp, Signal, and Telegram all tie identity to a phone number, which is itself a form of metadata that reveals who talks to whom, when, and how often.

The AI Frontier: Why the Stakes Are Even Higher

The shift away from US office software, video conferencing, and messaging is becoming broadly understood in European policy circles. What is less discussed, and more urgent, is what is happening in AI.

AI assistants represent a fundamentally different category of privacy risk compared to traditional software. When you use an open-source office suite, your document stays on your machine. When you use an AI assistant, everything you type, every question you ask, every document you upload, every business strategy you explore, flows to a remote server operated by a company in another jurisdiction.

And right now, those companies are deepening their ties to US defense and intelligence infrastructure in ways that Europeans need to understand clearly.

Claude.ai / Anthropic: Now Running on SpaceX Infrastructure

Anthropic, the company behind Claude.ai, announced in May 2026 a deal to use the full computing power of SpaceX’s Colossus 1 data center in Memphis, Tennessee, which houses more than 220,000 Nvidia processors and provides 300 megawatts of computing capacity. Anthropic has also expressed interest in working with SpaceX to develop orbiting AI data centers, taking European user data quite literally off the planet.

SpaceX is Elon Musk’s company. The data center is American infrastructure. The deal was announced on the same day that major US tech companies agreed to allow the US federal government access to test their AI tools. The Department of Defense simultaneously announced deals with ten different AI companies to use their technology on classified networks.

Anthropic has also separately partnered with Palantir and AWS to provide US intelligence and defense agencies with access to the Claude family of AI models. Claude is available in Palantir’s defense-certified environment, which handles data classified up to the “secret” level.

What this means for a European user: the AI assistant many people use for drafting business emails, summarizing sensitive documents, and working through strategic decisions is processed on infrastructure with direct ties to US defense and intelligence. That is not speculation or alarmism. That is what the companies themselves announced publicly this week.

Google Gemini: Aggressive Data Training, Even for Paying Users

Starting September 2, 2025, Google updated Gemini’s data usage policy to give itself authority to use “a sample” of users’ data to train its large language model. The data can be of any format, including documents, audio, videos, or images. Users who do not want their data used for training must manually opt out.

Gemini’s privacy policy explicitly states that Google “may use anonymized Personal Information to train AI models for internal use only.” On standard consumer-facing plans, user data is used for model training by default unless you actively navigate to settings and disable it which also disables your chat history.

The business-account exception is real but meaningfully narrow. The average individual European user, or small business owner, is not on a full Google Workspace Enterprise contract. They are on a consumer or small-business tier where the default is training on their data.

The broader pattern is consistent across US AI providers: if you are paying €20 per month for any of the major AI subscriptions, you are most likely still training their models by default. Paid individual plans typically require explicit user action to opt out of data training. Upgrading to a “Pro” account buys features, not privacy.

ChatGPT / OpenAI: Defense Contracts and US Government Access

OpenAI has been actively expanding its relationships with US defense and intelligence agencies. Department of Defense agreements announced in early 2026 include OpenAI products deployed on classified networks. ChatGPT conversations, business analyses, and personal questions asked by Europeans flow to US-based infrastructure that is increasingly integrated into American government operations.

The pattern across all major US AI providers is consistent: European users generate data that flows to the United States, where it is processed on infrastructure with defense contracts, subject to the CLOUD Act, and increasingly integrated into US government intelligence workflows. The conversations Europeans have with these AI assistants, their medical questions, legal concerns, business plans, and personal worries, are not staying in Europe.

CamoCopy: The European Alternative for AI

This is the gap that CamoCopy was built to fill.

CamoCopy is an AI assistant developed in Europe, running exclusively on European infrastructure, operated by European providers. Its architecture is built around a single principle that the major US providers have structurally abandoned: your conversations are your own.

Specifically:

Your data stays in the EU. Unlike US AI providers whose infrastructure routes data to American data centers, CamoCopy processes everything within European borders.

Your conversations are not used to train AI models. Unlike Gemini, which defaults to training on your data and requires manual opt-out, CamoCopy does not use your conversations for model improvement.

There is no CLOUD Act exposure. Because CamoCopy’s infrastructure is entirely European, there is no legal mechanism through which US authorities can demand access to your conversation data.

There are no defense contracts. CamoCopy has no ties to US intelligence or military infrastructure. Your business strategies and personal queries do not end up in environments with security clearance levels.

The choice of AI assistant is not a purely technical question. It is a question about where your most private thoughts, business strategies, legal questions, and sensitive documents actually go when you press send. For Europeans who have been following what their governments are doing about US office software, video conferencing, and messaging, the answer for AI should follow the same logic.

The Bigger Picture: A Continent Waking Up

What is happening across Europe is not a collection of isolated government IT decisions. It is a coordinated, accelerating shift in how European institutions think about digital infrastructure as strategic sovereignty.

The European Commission is soon presenting its “Tech Sovereignty Package”, which will include the Cloud and AI Development Act and measures aimed at restricting how non-European cloud providers handle sensitive government data in sectors including healthcare, finance, and judicial systems. A Commission spokesperson described the initiative as “Europe waking up and getting its act together.”

At KubeCon Europe 2026 in Amsterdam, the general manager of Linux Foundation Europe explained that while technologies like confidential computing can protect data technically, there is no technical answer if the US government insists on an American company flipping the kill switch on European email, office software, or access to US-hosted data. He stressed that this risk is no longer theoretical: it is happening, in public, with formal government announcements.

The question for European individuals and businesses is not whether their governments are moving in this direction. They clearly are. The question is whether private citizens and companies will wait for legislation to force change, or whether they will make these decisions now, while they still have the freedom to choose rather than comply.

Replacing your video calls with a European platform takes an afternoon. Switching your office suite to LibreOffice costs nothing. Moving your team messaging to Wire or Olvid takes an hour. And choosing a European AI assistant instead of feeding your most private conversations into an American system with defense contracts and CLOUD Act obligations is, increasingly, simply the rational next step.