Responsible AI
AI That Advises. Institutions That Decide.
AI on KRISHI360 is analytical and advisory — scoring, forecasting, flagging, benchmarking. It informs institutional decisions; it never makes regulated ones.
app.krishi360.in/audit
Access & AuditLive
Role checks today1.2M100% pass
Consent records8.4Lversioned
Incidents030 days
Institution B — portfolio viewRole-checked · purpose-boundAllowed
Research extract requestAggregation rules appliedScreened
Consent v3.2 — farmer 88412Recorded · revocableActive
Every access leaves a trace
Least privilege enforced
No automated regulated decisions
No automated credit sanctioning, no automated claim settlement, no automated benefit eligibility. Those decisions remain with the authorised institution, with a human in the chair.
Explainability
Signals carry their evidence: an institution can always ask why a flag was raised and see the underlying records.
Model governance
Documented model purpose, versioning, monitoring and review.
Bias monitoring
Ongoing evaluation to detect and correct systematic bias across regions, crops and farmer categories.
Data discipline
Models train and operate on authorised data under the platform's data-governance boundaries.
