Legal information
Artificial Intelligence Notice
Effective date: 7 June 2026
Human oversight: AI-supported outputs may be incomplete, biased or incorrect and should not be treated as final decisions without review.
1. Potential uses
AI may support research, document analysis, eligibility workflows, anomaly detection, communication, prototyping and system operations.
2. No fully automated high-impact decisions by default
Projects affecting access to essential assistance, identity or rights require appropriate human oversight, appeal and accountability mechanisms.
3. Data quality and bias
Models may reproduce bias or errors in source data. Testing, monitoring and documentation are required.
4. Confidentiality
Confidential, personal or security-sensitive information should not be entered into external AI services without approval and safeguards.
5. Project-specific governance
The legal basis, role of AI, human review, auditability and user notice are defined separately for each implementation.