Germany’s migration-AI proposal would make applicants’ data part of system development

The proposed law would let migration authorities use personal data to develop and test AI, monitor patterns in proceedings, and compare certain applicants’ statements with information on the internet.

Critics focused on how those powers would be bounded. A requirement that algorithms must not discriminate does not itself establish how discrimination will be detected. The federal data-protection commissioner also argued that a human checking the final result is insufficient if the decision to launch an intrusive search is itself automatic.

The government presented efficiency and security benefits; rights groups sought substantial revisions or abandonment. This is a record of a legislative proposal, not a description of settled law. Its lasting issue is where human judgment belongs: before a consequential data search begins, as well as after its output arrives.