On 29 July 2026, the cabinet led by Chancellor Friedrich Merz approved a draft law that creates a legal basis for the use of AI in the work of migration services. The document covers the processing of asylum applications, the issuance of visas, and the granting of residence permits.
The aim of the initiative is to speed up and standardise administrative procedures against the backdrop of persistently high application numbers and limited staffing. According to the government, AI is expected to improve the efficiency, quality and security of processes, as well as to unify practices nationwide.
What the bill specifically allows
Training and testing of AI systems. Authorities, including the Federal Office for Migration and Refugees (BAMF), local immigration offices and the Foreign Ministry, will be able to use personal data from real applications to train, validate and test artificial intelligence models. The systems will learn to identify relevant information, verify evidence and detect contradictions.
Analysis of already decided cases. AI will make it possible to generalise decision-making practices, identify typical patterns and look for opportunities to simplify procedures.
Automated comparison with open internet data. Where there are justified doubts about the accuracy of information (for example, contradictory details about origin, nationality, travel routes or the authenticity of documents), authorities will be able to cross-check the applicant’s information against publicly available sources online.
At the same time, the final decision in each individual case will remain solely with a human case officer. Algorithms will serve only in a supporting role.
After cabinet approval, the bill will proceed to the Bundestag and Bundesrat. Parliamentary hearings are expected in autumn 2026. The Interior Ministry emphasises that implementation will take place in compliance with data-protection and transparency requirements.
The initiative has drawn criticism from human-rights organisations, in particular PRO ASYL. They point to the risks of algorithmic bias, the use of inaccurate internet data, and possible pressure on staff who may uncritically accept AI suggestions under high workload.
The government insists that the law creates the necessary legal framework for the safe and privacy-compliant use of technologies that should relieve overloaded agencies and shorten processing times for applications.