German and EU Civil Society Warn Against Weakened AI Surveillance and Safety Rules

Civil society groups in Germany, including Amnesty International and the Chaos Computer Club, urged the government to withdraw draft laws that would expand digital policing powers through biometric internet searches and automated analysis of large police datasets using systems such as Palantir. Critics said the proposals from the justice and interior ministries lack judicial oversight, transparency, documentation requirements, and clear limits on data scope and analytical methods, creating risks of mass surveillance, discriminatory profiling, and intrusive scrutiny of victims, witnesses, and uninvolved people. Germany’s independent data protection authorities also concluded that the measures, as drafted, are incompatible with constitutional requirements and could effectively sidestep the EU AI Act’s ban on mass facial-image processing into biometric databases.

At the EU level, a coalition led by BEUC and 31 other organizations warned that the proposed AI Omnibus could dilute safeguards by exempting sectors such as medical devices, radio equipment, toys, and machinery from the AI regulation’s direct scope. The groups argued that existing sector-specific product rules do not address AI-specific harms including discrimination, opacity, and the evolving behavior of AI systems, and said the change would create regulatory gaps, fragmentation, and legal uncertainty rather than simplification. They warned that weakening the framework would undermine consumer protection, fundamental rights, and trust in European AI governance as trilogue negotiations continue.