Members of the European Parliament have rejected proposals that would permit tech companies to continue monitoring private messages.

The European Parliament [1] has rejected the proposed Chat Control measure, which aimed to implement automated scanning of private photos and text messages. The decision came down to a single vote, highlighting a narrow margin that prevented the ongoing mass surveillance of private communications by American technology firms. Following this decision, the remaining elements of the proposal did not garner the necessary majority support.

During the vote [2] in the chamber 311 Members of the European Parliament voted to oppose a motion to extend a derogation to the e-Privacy directive, while 228 members supported it, and 92 abstained. This decision means that technology companies are no longer legally permitted to conduct mass scanning of private messages.

However, law enforcement agencies retain the ability to surveil private messages if they have specific suspicions and have obtained a judicial warrant. They can also carry out routine scanning of public posts and files. The European Commission initially proposed a requirement for all email and messaging providers to implement mass scanning of all messages, including those that are end-to-end encrypted, in 2022. This proposal faced significant criticism from both technology companies and legal experts.

Automated AI assessments focused on images and videos that the algorithms had not encountered before. Text analysis examined private chat conversations for potentially suspicious language. This process occurred without a warrant, individual suspicion, or significant European oversight. The AI-driven examination of unfamiliar images and texts was determined to be flawed, according to [3] a recently published study by researchers at KU Leuven and Ghent University.

In 2024, the German [4] Federal Criminal Police Office has indicated that in 48% of all flagged chats were deemed criminally irrelevant. This suggests that nearly half of the data generated by the surveillance system consisted of inconsequential private conversations between individuals who were not involved in any illicit activity. Such a surge of misleading reports has strained investigative resources that could have been directed toward legitimate cases. The Federal Criminal Police Office [5] reported that 99,375 of the 205,728 reports forwarded by the National Center for Missing and Exploited Children were deemed not criminally relevant, resulting in an error rate of 48.3%. This figure marks an increase from 2023, when the number of false positives was recorded at 90,950.

Patrick Breyer, [6] who was a Member of the European Parliament, described outcome as [7] historic following his long-standing opposition to Chat Control. “This historic day brings tears of joy! The EU Parliament has buried Chat Control a massive, hard-fought victory for the unprecedented resistance of civil society and citizens! The fact that a single vote tipped the scales against the extremely error-prone text and image search shows: Every single vote in Parliament and every call from concerned citizens counted!,” said Breyer. “We have stopped a broken and illegal system. Once our investigators are no longer drowning in a flood of false and long-known suspicion reports from the US, resources will finally be freed up to hunt down organized abuse rings in a targeted and covert manner.”