AI systems
Register an AI System
Registration records enough operational context to classify the system and generate relevant governance work without repeatedly asking for known facts.
What to prepare
- Basic information and business purpose: what the system does, for whom, and who owns it.
- Provider and model: supplied services, models, and dependencies.
- Deployment: lifecycle stage, locations, jurisdictions, and intended use.
- Data and usage: input/output data, personal or sensitive data, scale, and autonomy.
- Affected people and human oversight: who may be affected and where people review or intervene.
- Review context: known risks, documentation, prior assessments, and relevant frameworks.
Why the information matters
- Jurisdiction helps determine which obligations may apply.
- Data types influence privacy and impact-assessment work.
- Provider or deployer role may alter responsibilities.
- Impact, autonomy, and affected people shape assurance and approval needs.
How to use it
- 1Open the Project's AI Systems area and choose New AI System.
- 2Complete the full-page registration sections using verified information.
- 3Save the record, identify gaps, and continue to classification.