The German Police Union in Berlin, the DPolG, is right on one fundamental point: a police force expected to confront organised crime, terrorism, cybercrime and increasingly mobile criminal networks cannot be forced to rely on the technological capabilities of a previous era. Criminal actors communicate digitally, exploit interconnected data environments and use automated tools themselves. It would therefore be unrealistic to deny law-enforcement agencies access to modern analytical technologies as a matter of principle. The debate surrounding Berlin’s amended General Security and Public Order Act, the ASOG, which came into force at the end of 2025, becomes too simplistic whenever every new digital police capability is portrayed automatically as a step towards a surveillance state.
Yet the DPolG’s response is equally incomplete when it dismisses expressions such as “super-database” or “biometric mass surveillance” primarily as political exaggeration. The Greens and the Left Party have brought a constitutional review application before Berlin’s Constitutional Court, and their challenge is directed not against the abstract idea of modern policing but against specific provisions with potentially far-reaching consequences. These include access to digital systems, automated data analysis, AI-related processing and the use of biometric technologies. Whether those powers are constitutionally proportionate is therefore not a rhetorical question but precisely the kind of issue a constitutional court is supposed to examine.
The biometric provisions illustrate why the discussion deserves greater precision. Under the amended ASOG, Berlin police may under certain circumstances conduct retrospective automated comparisons of facial and voice data against publicly accessible personal information available online. The rules can also extend in certain cases to contact persons and associates. Safeguards such as logging, oversight and data-protection controls exist, but that does not remove the fundamental significance of the capability itself. A system that can correlate biometric characteristics with large volumes of publicly available information is qualitatively different from conventional investigative work, even when its use is legally restricted.
The constitutional concerns are not limited to the political opposition. Berlin’s Commissioner for Data Protection and Freedom of Information has also raised substantial objections to several of the new investigative powers, particularly the scope of biometric internet searches and the use of lawfully stored police data for testing and training AI systems. The concern is especially relevant where analytical measures have a broad reach and can inevitably process data relating to people who are not themselves suspected of wrongdoing.
This is where the real conflict lies, and it is more important than the current political confrontation between the police union and the opposition suggests. The issue is not whether Berlin needs a technologically capable police force; given the development of digital crime, that proposition is difficult to dispute. The more difficult question is how far the state may aggregate data, analyse biometric characteristics and use personal information for new forms of algorithmic processing before a targeted investigative power becomes a structurally broad surveillance capability.
The security industry should pay close attention to this distinction because automated data analysis changes policing not only quantitatively but qualitatively. Investigators once had to connect information from separate cases and databases largely through manual work
