Back to Home
LangEval Legal

Privacy Policy

Version 2.1Last Updated: February 14, 2026

At LangEval, we believe that trust is the foundation of AI evaluation. This policy outlines our comprehensive approach to handling your data with the security and transparency required for enterprise deployments.


1Data Collection

We minimize data collection to what is strictly necessary for providing our evaluation services. When you use our SDK or platform, we may process:

  • Account Information: Name, email address, strictly for authentication and billing.
  • Evaluation Traces: Inputs and outputs of your AI Agents sent via our SDK. PII Masking is enabled by default to ensure sensitive customer data does not leave your premises in raw form.
  • Usage Telemetry: Anonymized metrics on SDK version, latency, and error rates to improve system stability.

2How We Use Data

Your evaluation data belongs to you. We do not train our internal models on your evaluation datasets or agent outputs. We strict use your data to:

  • Provide real-time scoring and observability dashboards.
  • Generate automated red-teaming reports.
  • Detect anomalies in your agent's performance.

3Data Sharing & Subprocessors

We do not sell your data. We share data only with trusted infrastructure providers (e.g., AWS, Cloudflare) bound by strict DPA (Data Processing Agreements). A full list of subprocessors is available upon request.

4Security Measures

We implement defense-in-depth security strategies:

Encryption

AES-256 encryption at rest and TLS 1.3 in transit.

Access Control

Strict role-based access (RBAC) and audit logging.

5Your Rights

Under GDPR, CCPA, and other privacy laws, you have the right to request access, correction, or deletion of your personal data. You can export all your evaluation data via our API at any time.

© 2026 LangEval Platform. All rights reserved.