Data practices · Last updated Jul 2026
Privacy Policy
Read as a ledger: each section opens with a one-line summary, then the section itself, in full. Where your text goes is set apart.
- Sections
- 11
- Read time
- 10 min
- Last updated
- Jul 2026
Legatus is committed to protecting your privacy. This policy explains what data we collect, how we use it, and the choices available to you.
Each section opens with a summary in navy. Summaries are orientation only and are not part of this policy — the section text governs.
Information We Collect
#Your account details, anything you save to a claim, and an audit log of what you did in the app. The claims workspace never leaves your browser, account or not.
Account information: If you create an account, we collect your name, email address, and a hashed version of your password. We never store passwords in plain text.
Claims data: Logged-in users may save claims, deadlines, saved analyses, and evidence files. This data is stored in our database and associated with your account.
Workspace data: The claims workspace keeps its data in your browser’s local storage (localStorage). It is never sent to our servers, whether or not you have an account, and it does not follow you to another device or browser.
Usage data: We maintain an audit log of actions taken within the platform (such as creating a claim or running an AI analysis) for security and feature-gate tracking purposes.
Text you type into an AI feature is sent to three named companies to be processed. None of them train on it, and on document features your personal identifiers are stripped out first.
When you use AI-powered features (chat, nexus evaluation, denial analysis, buddy statement generation, condition finder, hearing practice, document classification, or the rating criteria check), the text you submit is sent to third-party providers for processing:
Anthropic
Our AI model provider. Generates the educational responses for all AI features. Anthropic’s commercial API does not train on customer data.
OpenAI
Semantic-search (embeddings) provider only. Converts your search text and our regulatory reference library into numeric vectors so we can retrieve the most relevant regulations. OpenAI performs no generative processing on your data and does not train on it via the commercial API.
Langfuse
AI observability provider. Receives a record of each AI request (prompt content and the model’s response, token counts, and a one-way-hashed user identifier) so we can monitor quality, cost, and errors. See the Sub-Processors section below for details.
Redacted on document features
SSN · date of birth · address · phone · email · VA file number
Text submitted to AI features is processed in real time and is not used to train AI models. For document-based features, personal identifiers (Social Security numbers, dates of birth, addresses, phone numbers, email addresses, and VA file numbers) are automatically redacted before your text is transmitted to any AI provider or observability service. Refer to Anthropic’s Privacy Policy, OpenAI’s Privacy Policy, and Langfuse’s Privacy Policy for details on how they handle data sent to their services.
File Storage
#Uploaded evidence is stored privately and is reachable only from your account.
Files uploaded to the Evidence Vault are stored securely via Vercel Blob. Files are associated with your account and are accessible only to you. Accepted file types include PDFs, images, DOCX, and plain text files, subject to tier-based storage limits.
Authentication & Sessions
#A session cookie keeps you signed in. No advertising, tracking or analytics cookies.
Authentication is handled via NextAuth.js using a JWT (JSON Web Token) strategy. We use session cookies to maintain your logged-in state. We do not use tracking cookies, advertising cookies, or third-party analytics cookies.
Third-Party Services
#Beyond the AI providers, other companies run parts of the platform: payments, hosting, the database and email.
In addition to the AI providers listed above, we use the following third-party services:
Stripe
Payment processing for subscription billing. Stripe handles all payment card data directly; we never see or store your full card number. See Stripe’s Privacy Policy.
Vercel
Application hosting and file storage (Vercel Blob).
Neon
PostgreSQL database hosting.
Resend
Email delivery — password resets, magic link authentication, and the claim deadline reminders described below.
Claim deadline reminders. If you use the Navigator, we email you when a deadline recorded on your claim is approaching — an intent-to-file window, an appeal window, a response VA is waiting on, a proposed rating reduction, or a records request that has gone unanswered. These emails contain the deadline date and the name of the record or issue it belongs to. They are not marketing, and we do not send them to anyone who is not running a claim in the Navigator.
They are on by default, because a missed VA deadline can cost you money or a claim. You can turn them off at any time in Settings, or with the unsubscribe link at the bottom of every one of these emails — that link needs no login. Turning them off stops the reminders only; it does not affect password resets or sign-in emails, which you need in order to use your account.
Data Retention & Deletion
#We keep your data while your account is open. Deleting the account is immediate and you can do it yourself. Logs and AI traces expire on their own schedules.
Your account data, claims, deadlines, saved analyses, and evidence files are retained for as long as your account is active. If you downgrade to a free tier, excess evidence files may be archived but are not immediately deleted.
Retention schedule:
- Verification tokens
- purged automatically after expiration
- Security audit logs
- retained for 2 years, then purged
- AI usage metrics
- retained for 1 year, then purged
- AI observability traces (Langfuse)
- pseudonymized (identified only by a one-way hash that cannot be linked to you without an internal secret) and retained for up to 30 days (our current Langfuse plan’s data-retention window), then purged
You can delete your account and all associated data directly from Settings > Data & Privacy. Account deletion is processed immediately and removes your profile, claims, evidence files, saved analyses, and associated data. AI observability traces are pseudonymized and cannot be linked back to you; where a linkable-deletion obligation applies, we will honor it using the provider’s deletion tools. You can also email support@legatusos.com.
Your Privacy Rights
#California and EU/UK residents have specific, named rights — and everyone, everywhere, can export or delete everything from Settings without asking us.
For California Residents (CCPA/CPRA)
Under the California Consumer Privacy Act and the California Privacy Rights Act, California residents have the following rights:
- Right to Know — request the categories of personal information we have collected, the purposes for collection, and the categories of third parties with whom we share it
- Right to Delete — request deletion of your personal information (available in Settings > Data & Privacy or by emailing us)
- Right to Correct — request correction of inaccurate personal information
- Right to Opt-Out of Sale/Sharing — we do not sell or share your personal information for cross-context behavioral advertising
- Right to Non-Discrimination — exercising your rights will not result in denial of service or different pricing
- Right to Limit Sensitive Personal Information Use — disability status and health-related claim data are sensitive personal information under CPRA and are used only to provide the Service
Categories of personal information collected: identifiers (name, email), health-adjacent information (claimed conditions, evidence descriptions, nexus letter text), usage data (audit logs, AI feature usage counts), and payment information (processed by Stripe; we do not store card numbers).
How to submit requests: use the self-service tools in Settings > Data & Privacy or email support@legatusos.com. We will verify your identity and respond within 45 days.
For EU/UK Residents (GDPR)
Under the General Data Protection Regulation, EU and UK residents have the following rights:
- Right of Access (Art. 15) — obtain a copy of your personal data
- Right to Rectification (Art. 16) — correct inaccurate data
- Right to Erasure (Art. 17) — request deletion of your data
- Right to Restriction (Art. 18) — limit processing of your data
- Right to Data Portability (Art. 20) — export your data in a machine-readable format (available in Settings > Data & Privacy)
- Right to Object (Art. 21) — object to processing based on legitimate interests
Legal bases for processing: contract performance (providing the Service you signed up for), legitimate interests (security, fraud prevention, product improvement), and consent (where specifically obtained).
Data controller: Legatus. Contact: support@legatusos.com.
International data transfers: your data is processed in the United States by our service providers (Anthropic, OpenAI, Vercel, Neon, Resend, Langfuse). These transfers are governed by standard contractual clauses or equivalent mechanisms. You have the right to lodge a complaint with your local supervisory authority.
For All Users
Regardless of your location, you can export your data or delete your account at any time from Settings > Data & Privacy. Data export provides a JSON download of your profile, claims, assessments, evidence metadata, saved outputs, deadlines, subscription details, and audit log.
Sub-Processors
#The full list of companies that touch your data, what each one receives, and where they are. Every one of them is in the United States.
We use the following sub-processors to provide the Service. For each, we describe the categories of data transmitted:
Anthropic
United States
AI model provider; generates the educational output for all AI features. Data sent: user-submitted text for AI features (claim descriptions, questions, buddy statement answers, evidence text extracts). On document-derived paths, personally identifiable information (SSNs, dates of birth, addresses, phone numbers, email addresses, VA file numbers) is redacted before transmission. Anthropic’s commercial API does not train on customer data.
OpenAI
United States
Embeddings / semantic-search provider only. Data sent: the text of your search queries and questions, converted into numeric vectors used to retrieve relevant regulations. On document-derived paths, personally identifiable information is redacted before transmission. OpenAI performs no generative processing on this data and does not train on it via the commercial API.
Langfuse
ClickHouse, Inc. · United States
AI observability and quality monitoring. Data sent: a trace of each AI request — the prompt content and the model’s response, token-usage counts, model and feature identifiers, and a pseudonymous user identifier (a one-way SHA-256 hash that cannot be linked back to you without an internal secret). On document-derived paths, personal identifiers are redacted before the trace is created. Traces are used only to monitor AI quality, cost, and errors, are never used to train models, and are retained for up to 30 days.
Vercel
United States
Application hosting and file storage (Vercel Blob). Data sent: uploaded evidence files (PDFs, images, documents), application logs.
Neon
United States
PostgreSQL database hosting. Data stored: account information, claims data, evidence metadata, audit logs, subscription records. Sensitive extracted text is encrypted at rest (AES-256-GCM).
Stripe
United States
Payment processing. Data sent: email address for customer identification. Stripe handles all payment card data directly; we never see or store card numbers.
Resend
United States
Transactional email delivery. Data sent: email address and email content (password resets, verification links, consent notifications).
Children’s Privacy
#Legatus is for adults. You must be 18 to use it.
Legatus is not intended for use by individuals under the age of 18. We do not knowingly collect personal information from children.
Changes to This Policy
#We can change this policy. Material changes come by email or an in-app notice, and the date at the top changes with them.
We may update this privacy policy from time to time. Material changes will be communicated via email or an in-app notice. Changes will also be reflected by updating the “Last updated” date at the top of this page. Continued use of the platform after changes are posted constitutes acceptance of the updated policy.
Contact
#Questions about your data go to support — a person reads them.
If you have questions about this privacy policy or how your data is handled, contact us at support@legatusos.com or visit our Contact page.
Version history
3 REVISIONS- Jul 2026Named Langfuse as an AI observability sub-processor, and described the personal identifiers redacted before a trace is created.
- Mar 2026Added the CCPA/CPRA and GDPR rights sections, the sub-processor list with the data sent to each, the retention schedule, and how sessions are handled.
- Feb 2026First publication.
Questions about your data? support@legatusos.com
