Privacy Policy
Last updated: October 2026
1. Our Core Privacy Philosophy
At SecureMID ("SecureMID", "we", "us"), we believe that an AI security platform must maintain an uncompromised privacy standard. We do not transmit customer model inputs, output completions, or vulnerability traces to external commercial cloud APIs by default.
2. Information We Collect
- Account & Organization Information: Name, work email address, company name, and encrypted authentication credentials.
- AI Agent Configuration: Endpoint URLs, authentication headers, and environment labels configured in the Asset Registry.
- Assessment Traces & Evidence: Conversational turns, tool invocations, and assertion evaluation outcomes generated during authorized red teaming assessments.
3. Air-Gapped Local Judge & Zero Data Egress
Security assertion evaluations are performed by fine-tuned models running in-process or within the customer's isolated security perimeter. The setting SECUREMID_ALLOW_CLOUD_FALLBACK=false is enforced by default to guarantee zero outbound transmission of assessment traces.
4. Data Retention & Pruning
Assessment records and cryptographic audit traces are retained in accordance with your subscription plan tier:
- Free Community: 7 days retention.
- Pro: 90 days retention.
- Enterprise: Tailored contract retention with automated cryptographic pruning upon request.
5. Contact Our Privacy Officer
For privacy inquiries, data deletion requests, or compliance questionnaires, contact us at security@securemid.com.