top of page

Data Usage & AI Security Disclosure

Last Updated: February 10, 2026

At Asthra Writer, we prioritize the security and confidentiality of your intellectual property. Our AI-powered drafting assistant is purpose-built for the life sciences sector, where data integrity and privacy are non-negotiable. This disclosure explains how we handle your document content and the measures we take to ensure your data remains secure.

1.Data Encryption in Transit

Your data is protected from the moment it leaves Microsoft Word.

  • HTTPS Protocol: All communication between the Asthra Writer add-in and our processing servers is conducted exclusively over industry-standard HTTPS (TLS 1.2+) encryption.

  • End-to-End Security: This ensures that document content, source materials, and AI-generated drafts cannot be intercepted or read by unauthorized third parties during transmission.

2. Session-Based Architecture (No Data Retention)

We follow a strict "Zero Persistence" policy for all user-provided document content.

  • In-Memory Processing: When you select text or upload a source document, the content is processed in volatile memory (RAM) solely for the duration of your active drafting session.

  • No Permanent Storage: We do not store, log, or cache your document content, tables, or images on our permanent storage servers once your session is concluded.

  • Immediate Disposal: Once the AI has generated the requested draft and the session ends, all temporary data used for that specific task is discarded.

3. AI Usage and Data Governance

Asthra Writer uses a hybrid stack of frontier and domain-tuned Large Language Models (LLMs) to identify relevant sections and create evidence-backed drafts.

  • Purpose of Processing: Content is transmitted to our AI engines for the sole purpose of generating structured regulatory drafts, such as CSRs, PSURs, or CTD Modules.

  • No Model Training: We have strict agreements with our AI infrastructure providers to ensure that your sensitive document content is never used to train public or foundational AI models.

  • Human-Free Workflow: No humans at Asthra AI or our service providers have access to view, review, or manually label your document content.

4. Application Permissions

To function within Microsoft Word, our add-in requires specific permissions that are regularly audited for compliance.

  • Read/Write Access: The add-in uses the ReadWriteDocument permission level to read your source context and insert generated drafts directly into your document.

  • Scope of Access: The add-in only accesses the specific document you are currently working on and does not have visibility into other files on your local machine or cloud storage.

5. Transparency and Compliance Review

We design our systems to be transparent for internal organizational reviews and external certifications.

  • Audit Ready: Our architecture is designed to meet the rigorous privacy standards required for life sciences documentation.

  • Clear Disclosures: We provide this disclosure to meet Microsoft AppSource certification requirements and to give our users full visibility into our data practices .

Contact Information

For questions regarding these Terms or for technical support, please visit our support page or contact us at:

Email: info@asthra-writer.ai

Website: www.asthra-writer.ai

bottom of page