AI in German Migration Administration: Concerns Over Draft Legislation

• The German government is pursuing legislation to integrate AI into asylum and visa application processes to improve efficiency.
• Key provisions include training AI on personal data, automated procedure monitoring, and automated cross-referencing of online data for verification.
• Critics, including human rights organizations and data protection agencies, warn of risks regarding discrimination, privacy, and technical errors.

Key Legislative Changes

• Data Usage: Agencies managing asylum and residence laws will be permitted to train and test AI systems using personal data.
• Automated Monitoring: Implementation of systems to track procedural trends, such as identifying if certain applicant profiles are rejected at higher rates.
• Online Verification: Authorities may automatically cross-reference applicant data with publicly available internet information to verify claims.
• Data Protection: The final draft explicitly excludes biometric data from these automated checks.

Criticisms and Concerns

• Insufficient Anti-Discrimination Measures: Federal Anti-Discrimination Commissioner Ferda Ataman and civil society groups argue that mere requirements to avoid "discriminatory algorithms" are insufficient; they demand transparency reporting and rigorous auditing.
• Fundamental Rights: The Federal Commissioner for Data Protection and Freedom of Information (BfDI) warns that efficiency must not undermine constitutional principles. They argue for strict hierarchies in data usage (anonymized vs. personal) and object to automated triggering of data checks.
• Human Rights Impact: Organizations like 'Pro Asyl' describe the bill as a "blank check" for state agencies, emphasizing that these systems handle decisions affecting human lives.

International Precedents and Risks

• Netherlands: Investigative reports revealed that an automated visa classification system created discriminatory outcomes, unfairly flagging individuals based on nationality and age.
• United Kingdom: A judge criticized the UK Home Office for relying on "AI-hallucinated" evidence in an asylum denial, where the ministry cited a document that did not exist.
• Technical Limitations: Critics fear that new age-recognition technologies in the UK could lead to more unaccompanied minors being incorrectly classified as adults.