Plain-language questions and answers
What this system does - and never does.
These answers describe the current synthetic engineering demonstrator. They are not claims of government approval, production certification or real-court use.
Authority
Is this an AI judge?
No. It cannot issue judgment, choose a judge, recommend bail or sentence, score credibility, or decide a legal outcome. Authorised people remain responsible for every legal decision.
Data
Does the demonstration need court data?
No. It uses invented scenarios. Real data is blocked until a competent institution clears the documented legal, security, privacy, accessibility, identity, integration and operational gates.
Security
Is it production-certified?
No. Internal controls and tests exist, but institutional keys, identity services, independent assessment and production infrastructure are not connected or approved.
Configuration
Can one engine support different countries?
The demonstrator loads separate synthetic jurisdiction packs over one country-neutral core. An official pack would have to be mapped, reviewed and signed by the competent institution.
Pilot
What can an institution safely do first?
Run a synthetic, non-authoritative discovery exercise: map one narrow workflow, review controls, measure a baseline and decide whether further evaluation is justified.
Limits
What remains outside the builder's control?
Official workflow approval, government identity and keys, infrastructure, independent assurance, procurement, real pilot evidence and legal authorisation must come from competent external parties.