Trust & Safety
Trust is built in, not bolted on.
How YAS protects data, governs models, and uses AI responsibly. AURA supports underwriting decisions with explainable scores — the licensed insurer retains authority, and a qualified person makes the final call.
Three pillars
Security, responsible AI, governance.
Every YAS surface rests on the same three commitments. They are architectural, not promised in a contract.
Security
Enterprise-grade controls protect telemetry ingestion, scoring pipelines, and operator dashboards from edge to attestation.
Responsible AI
AURA is decision-support, not autonomous underwriting. Scores are explainable and monitored; humans retain authority.
Governance
Versioned model parameters, board-visible reporting, and carrier co-sign where programmes require it.
Security
Controls across the pipeline.
Encryption
Data is encrypted in transit and at rest across production systems.
Access control
Role-scoped access; requests are bounded to the fleet they're authorised for.
Audit trails
Scoring inputs, outputs, and explanations are logged and time-stamped.
Vendor review
Sub-processors are reviewed; material changes carry notice to customers.
Responsible AI
Decision-support, with a human in the loop.
AURA informs underwriting; it does not replace it. These commitments keep authority where it belongs.
Every score has a reason
Each score comes with a plain-language explanation of the behaviours behind it — readable by operators, decision-makers, and regulators alike.
A person makes the call
YAS provides the risk picture; it never binds coverage. A qualified person reviews and signs off every decision.
Changes are tested first
Before any model update goes live, it is tested against real operating data and must clear strict safety checks.
Fairness-tested
Scores are checked so that no group is treated unfairly, with monitoring for drift over time.
Governance
Accountable by design.
YAS maintains versioned model parameters and board-visible reporting, with carrier co-sign on attestation records where a programme requires it. Regulators expect AI to be open, accountable, and easy to explain — YAS was built that way from day one, not patched on later to pass an audit.
- Versioned model parameters with change history
- Board- and regulator-ready governance exports
- Carrier co-sign on attestation where programmes require it
- Documented human-review paths for material changes
Need our security documentation?
For security questionnaires, our DPA, or a governance walkthrough, contact us via the enterprise form.