Skip to content

Trust Center

Security, compliance, and contractual posture in one place. Built for the enterprise security reviews regulated life sciences teams run before procurement.

Asthra AI is designed for regulated life sciences environments. The short version: closed-system retrieval (no open-internet, no training-memory leakage), workspace isolation per customer, encryption in transit and at rest, SOC 2 Type 1 ready, and a defined path to SOC 2 Type 2, ISO 27001, GDPR, and HIPAA. Customer data is never used to train models — ours or Anthropic's. The detail lives below.

Compliance status

Honest about where we are and where we're going

SOC 2 Type 1

Controls over security, availability, processing integrity, confidentiality, and privacy are designed and operating. Report available on request under NDA.

Ready

SOC 2 Type 2

Extended testing period for continuous control effectiveness. Audit in progress; estimated completion aligned with customer onboarding timelines.

In progress

ISO 27001

Information security management system assessment scheduled. Gap analysis complete.

In progress

GDPR

Framework in place for EU personal data. Standard contractual clauses available in the DPA.

In progress

HIPAA

BAA available for US healthcare customers handling protected health information. Controls implemented.

In progress

Artifacts (SOC 2 letter, ISO 27001 scope, etc.) available under NDA — request via info@asthra-writer.ai.

Subprocessors

Third parties that may process customer data on our behalf

SubprocessorPurposeRegion
Anthropic (Claude API)Language model for agentic retrieval and drafting.United States
Cloud hosting (customer-selected)Compute, storage, and networking for the Asthra backend. Customer can choose AWS, Azure, or GCP for VPC deployments.Selectable per deployment
Observability / loggingApplication and infrastructure monitoring, audit log storage.Same region as deployment
Email deliveryTransactional email (deployment notifications, demo requests).United States

Material changes to the subprocessor list are notified to active customers in advance, per the DPA. For the most current list at any time, email info@asthra-writer.ai.

Data retention and deletion

Customer documents. Source documents and generated outputs are retained for the duration of your subscription. They are never used to train models — ours or a third party's.

Deletion on request. Honoured within thirty (30) days of written request, confirmed by email. Tighter SLAs can be agreed in the master services agreement.

Audit ledger. Transaction ledger entries required for regulatory traceability may be retained longer than the underlying source documents. The exact window is tied to your validation posture and agreed in the master services agreement.

Backups. Operational backups follow industry-standard retention, typically 30 days, and are included in any deletion request.

Security team

For security reviews, DPA negotiation, vendor onboarding, or any question that should not go through a generic contact form — reach us directly.

info@asthra-writer.ai

Last updated: 16 April 2026