# Introducing Conduktor Trust: in-stream data quality enforcement

Bad data (missing fields, broken formats, out-of-range values) breaks downstream systems, corrupts AI outputs, and causes customer-facing issues. Traditional tools fix data after the damage. Data contracts push responsibility to producers with fragmented enforcement.

**Conduktor Trust** enforces data quality in-stream, before bad data reaches your pipelines.

## How Trust works

1. **Define rules**: Use CEL (Common Expression Language) to specify structure, completeness, and conformance
2. **Apply policies**: Attach rules to specific topics and control where enforcement applies
3. **Act on violations**: Log violations for monitoring or block bad messages instantly
4. **Track patterns**: View violations over time to identify recurring issues

## What Trust prevents

- **Data quality issues reaching downstream**: Catch problems at the source
- **Bad data in AI models**: Feed only trusted data to models and analytics
- **Compliance gaps**: Apply consistent quality policies across teams
- **Reactive cleanup work**: Fix issues before they spread

---

For a full list of changes, read the [complete release notes](https://docs.conduktor.io/changelog/#console-1340).