The Battle Over Chat Control: How EU Governments and the Tech Lobby Are Trying to Overturn Parliament’s Vote — A Comprehensive Fact Check


The media in this post is not displayed to visitors. To view it, please log in.

This week, the European Parliament faces a decisive vote on whether the indiscriminate scanning of private chats and emails by US tech companies (Chat Control 1.0) will be allowed to continue. After Parliament voted on 11 March to replace blanket mass surveillance with targeted monitoring of suspects — thereby protecting the confidentiality of digital correspondence — EU member state governments let the trilogue negotiations fail by refusing to compromise in substance.

Now, in an unprecedented manoeuvre, the conservative EPP group is attempting to force a repeat vote on Thursday (26 March) to overturn the Parliament’s principled decision and keep indiscriminate chat scanning in place. A preliminary vote on Wednesday will determine whether this repeat vote goes ahead or is struck from the agenda.

Digital rights expert and former MEP Patrick Breyer outlines the urgently needed change of strategy:

“Indiscriminate Chat Control is like trying to mop up water while the faucet is still running. It is technologically obsolete and a proven failure in criminal justice terms. Flooding our police forces each year with hundreds of thousands of hits from unreliable US algorithms — most of them either false positives or long-known duplicates — does not rescue a single child from ongoing abuse. This data deluge ties up massive resources that are desperately needed for undercover investigations into actual abuse networks. To genuinely protect children online, we need a paradigm shift: providers must be required to prevent cybergrooming through safe app design and strict default settings. Illegal material on the open internet and darknet must be proactively tracked down and removed at source. That is what truly protects children.”

Background: What exactly expires on 3 April

An EU interim regulation (2021/1232), set to expire on 3 April, currently permits US corporations such as Meta to carry out indiscriminate mass scanning of private messages on a voluntary basis. Three types of chat control are authorised: scanning for already known images and videos (so-called hash scanning, which generates over 90% of reports); automated assessment of previously unknown images and videos; and automated analysis of text content in private chats.

The AI-based analysis of unknown images and texts is extremely error-prone. But the indiscriminate mass scanning for known material — proposed by socialists and liberals — is highly controversial, too: beyond the unreliability of the algorithms documented by researchers, these scans rely on opaque foreign databases rather than European criminal law. The algorithms are blind to context and lack of criminal intent (e.g. consensual sexting between teenagers). As a result, vast numbers of private but criminally irrelevant chats are exposed.

In the run-up to the vote, US tech corporations, foreign-funded lobby groups, and law enforcement agencies are flooding public discourse with warnings about an alleged “legal gap.” A comparison of their claims with internal documents, scientific studies, and the voices of child protection experts and actual abuse survivors, however, reveals an entirely different picture.


Disinformation Narratives of Chat Control Proponents — and the Facts


Disinformation 1: “The European Parliament is to blame for the collapse of negotiations and is putting children at risk.”
(Claimed by the lobby alliance ECLAG and US tech companies)

  • Fact: It was the EU Council of Ministers that deliberately let the trilogue negotiations fail, for tactical reasons.
  • Evidence: Leaked Council cables, classified as restricted, reveal that EU member states showed no willingness to compromise, fearing that any concession could set a precedent for the permanent Chat Control 2.0 regulation. The classified minutes from 13 March show that the Cypriot Presidency already anticipated failure before the final trilogue round, noting it did “not expect to reach an agreement” with the lack of a new mandate given by member states. A majority including Hungary, Belgium, Sweden, Spain, Latvia, Slovakia, Malta, Estonia, Slovenia, Romania and Germany were unwilling to make any concessions on scope. Only a small minority of governments including France and Ireland agreed to the Presidency’s proposal to phase out at least the most error-prone text scanning in search of “grooming”. The Netherlands showed itself “completely flexible”, and Italy had long before criticised the scope of scans and demanded “prior authorization of detection activities by [public] authorities”.
  • Parliament’s lead negotiator, Birgit Sippel (S&D), sharply criticised the Council after the breakdown: “with their lack of flexibility, Member States have deliberately accepted that the interim regulation will expire in April.”

Disinformation 2: “Without indiscriminate Chat Control, law enforcement will be flying blind.”
(Claimed by law enforcement officials across the EU)

  • Fact: Targeted telecommunications surveillance based on concrete suspicion and a judicial warrant remains fully available after 3 April, as does the bulk scanning of public posts and hosted files. User reports also remain possible. The real problem for authorities is a flood of false leads and a systemic refusal to remove material from the internet.
  • Evidence — investigative chaos: According to Germany’s Federal Criminal Police Office (BKA), nearly 50% of chat control reports are criminally irrelevant. This flood of data waste ties up massive resources desperately needed for targeted, undercover investigations into real abuse networks. Where investigations are opened, German crime statistics show that around 40% of suspects are minors themselves, often acting without criminal intent or in consensual situations. The Federation of German Criminal Investigators (BDK) warns that this mass surveillance produces “a flood of tips… often without any actual investigative lead.” Meanwhile, Europol and German authorities systematically refuse to proactively have abuse material removed from the internet, as investigative reporting by ARD/STRG_F has revealed — images and videos remain online despite authorities being fully able to have them taken down, even as they demand ever more surveillance powers.
  • Evidence — failure to protect children: Mass scanning for already known images does not stop ongoing abuse and does not rescue children in acute danger. According to the European Commission’s own evaluation report, no measurable link can be established between the mass surveillance of private messages and actual convictions. Yet the Commission and Council demand the extension of a measure whose effectiveness they themselves cannot demonstrate.
  • Evidence — risk of annulment in court: The European Data Protection Supervisor (EDPS) stresses that any solution used to detect illegal content must be targeted and not indiscriminate. The Council’s own legal service concluded in 2023 concerning the proposal for a permanent regulation (CSAR): “the detection order regime provided for by the proposed Regulation as regards interpersonal communications entails a serious risk that it would be found to compromise the essence of the rights to privacy and data protection enshrined in Article 7 and 8 of the Charter, in so far as it would seek to authorise access on a generalised basis, through automated and systemic screening surveillance, to the content of electronic communications and personal data of all users of a specific service, irrespective of their direct or indirect link with child sexual abuse criminal activities” (para 58)

Disinformation 3: “The scanning technology deployed is highly precise and protects privacy.”
(Claimed by Meta, Google, Microsoft, Snap, TikTok)

  • Fact: The technology is an ineffective legacy system, error-prone, and destructive to the security of private communications.
  • Evidence — an obsolete model: Offenders can effortlessly switch to secure messengers where no chat control takes place. Due to the increasing adoption of end-to-end encryption by providers, the number of chats reported to police has already dropped by 50% since 2022. Most recently, only 36% of reports from US companies originated from the chat control of private messages, while social media platforms and cloud storage services are becoming increasingly relevant. Rather than investing in targeted investigative work, the Council clings to a dying surveillance model.
  • Evidence — unreliability: A recent international research paper documents the structural weaknesses of the industry standard PhotoDNA. The software is unreliable: criminals can make illegal images invisible through minimal alterations (e.g. adding a border), while innocent citizens can easily be falsely flagged. In a November 2025 open letter, leading IT researchers (including from the universities of Aarhus, Leuven, and ETH Zurich) warned: “False positives seem unavoidable.” According to an open letter by a coalition of more than 40 civil liberties organisations and professional associations (including Europe’s leading digital rights groups), the Commission’s own evaluation report confirms the measure’s failure: the US algorithms deployed show error rates of 13 to 20 percent. Of the billions of messages scanned, only 0.0000027 percent were actually illegal material.

Disinformation 4: “The call for Chat Control comes primarily from victims and civil society.”
(Suggested by the ECLAG campaign)

  • Fact: Actual survivors are taking legal action against Chat Control. The real driving force behind the campaign is a network of tech companies and lobby organisations funded by governments and non-European foundations.
  • Evidence — survivors speak out: Survivors of sexualised violence are fighting back. Alexander Hanff, a survivor and privacy advocate, writes: “As a survivor, I depend on confidential communication to find support and report crimes. Taking away our right to privacy means further harming us.” Dorothée Hahne of the survivors’ association MOGIS e.V. warns: “We see our safe spaces destroyed.” To preserve safe spaces for victims, a survivor from Bavaria is currently suing with the support of the Society for Civil Rights (GFF) against Meta’s scanning of his chats. The civil society coalition also warns that indiscriminate scanning dangerously undermines professional confidentiality for lawyers, doctors, and therapists.
  • Evidence — lobbying: Who truly benefits from this legislation was exposed in an investigative report by Balkan Insight. The US organisation Thorn, which sells scanning software to public authorities, invests hundreds of thousands of euros annually in EU lobbying. ECLAG members are supported by tech corporations and the non-European Oak Foundation.

The Alternative to Surveillance Overreach: “Security by Design”


The European Parliament advocates a genuine paradigm shift, supported by civil society, survivor networks, and IT security experts: instead of indiscriminate mass surveillance of private communications using error-prone US algorithms, chat and messaging services should be “Secure by Design.” This includes:

  1. Strict default settings and protective mechanisms (Security by Design) to make cybergrooming technically harder from the outset and prevent the creation of CSAM.
  2. Targeted telecommunications surveillance based on judicially confirmed suspicion.
  3. Proactive search by a new EU Center and immediate takedown obligations for providers and law enforcement on the open internet and darknet — removing illegal material at source.

Call to Action

Civil liberties advocates are urging citizens across Europe to contact their MEPs directly ahead of the decisive votes on Wednesday and Thursday. Through the campaign page fightchatcontrol.eu, MEPs can be called upon to reject the undemocratic motion for a repeat vote and to uphold the fundamental right to confidential correspondence.

Breyer warns:

“When a democratic decision is put to a vote repeatedly until the desired outcome is achieved, Parliament itself is devalued. This approach sets a dangerous precedent. It undermines the reliability of democratic processes and sends the signal that majorities only count when they are politically convenient. The responsible actors are damaging not only trust in the European institutions, but the very foundations of democracy.”

On Tuesday, EU governments will strategise in a restricted format and behind closed doors on the issue.

Citizens can contact their representatives via: fightchatcontrol.eu


patrick-breyer.de/en/the-battl…


🇪🇺 1/7 🌍 Foreign-funded lobby groups from outside the EU are pushing #ChatControl with misleading propaganda. They want to #PassTheLaw to scan your chats, but who are they and who's paying them? Let's expose the network.
Thread 👇