Privacy Policy

Overview

Codex is an AI-first translation editor built by Frontier R&D Ltd. This policy explains what data we collect, how it is used, and the protections we put in place to keep your translation work private and secure.

AI Model Training and Your Data

Codex accesses AI capabilities through the APIs of leading model providers. For all paid plans, we use API-tier access under which customer data is explicitly excluded from model training. On Organization and Enterprise plans, your translation data, project files, and AI interactions are not used to train any AI model — by us or by our model providers.

Your project's terminology, style standards, and translation history are read on the fly to inform every AI interaction — giving you the benefit of a model that feels tailored to your work, without your data ever becoming part of the model itself. Your expertise stays in your project, not in a shared model.

Enterprise customers with dedicated deployments and custom data processing agreements receive additional contractual guarantees beyond these baseline commitments. See the Enterprise section below.

Where Your Project Data Lives

Your translation projects are stored on private, access-controlled infrastructure operated by Frontier R&D. Your project storage is:

• Private by default — not publicly accessible.
• Accessible only to members you explicitly invite, and to the Frontier R&D team for support and operations.
• Stored on servers we control; your data is not sold or shared with third parties.

AI inference requests are proxied through our own API layer before reaching any model provider. These requests contain only the context needed for the current interaction; we do not persistently log the content of AI inputs. We reserve the right to monitor API traffic for abuse, exploits, or troubleshooting.

Account and Authentication Data

When you register for Codex, we collect:

• Email address — used for login, account recovery, and transactional communications.
• Password — stored as a secure one-way hash; we never store or transmit your plaintext password.
• Subscription and billing data — handled by our payment processor. We do not store full credit card numbers.

Usage Analytics

We collect aggregated usage metrics — such as feature engagement and general usage patterns — to understand how Codex is being used and to improve the product. We do not include the text of your translation work in these metrics.

Desktop application (Codex Translation Editor): A/B testing analytics may send which variant you selected for product experiments, together with one-way cryptographic hashes derived from your account identifier and project name so we can correlate events without storing your email, display name, or project title in readable form. You can disable A/B testing in the extension settings when that option is available.

Website (codexeditor.app): We use privacy-oriented analytics (Umami) to measure page views and referrers. Umami does not use cookies for tracking and does not follow you across unrelated sites.

Enterprise: Dedicated Private Cloud Deployments

For organizations with heightened data-sovereignty or compliance requirements, Frontier R&D offers dedicated single-tenant deployments. In this configuration:

• Your organization's data is hosted in a fully isolated environment, completely separated from other tenants at the infrastructure level.
• No data crosses shared compute or storage resources with other organizations.
• The deployment can be configured for a specific geographic region to meet data-residency requirements.
• Access controls, audit logs, and network policies are scoped exclusively to your organization.

Enterprise agreements include negotiated data processing addenda that supersede the baseline terms in this policy where more restrictive protections are required. To discuss a dedicated deployment, contact us at admin@frontierrnd.com.

Your Content, Your License

You own your translation data. Organizations can attach a license to their project — such as a full copyright reservation or a Creative Commons license — and that license governs how contributors and downstream users may use that content. Frontier R&D does not claim any ownership or rights over content you create in Codex.

Subprocessors

Codex relies on a small number of third-party services to operate. We select subprocessors based in part on their data handling commitments, and we require that AI providers be used via API tiers that exclude customer data from model training.

We may update this list as subprocessors change. Material additions will be reflected here and noted in policy update notices.

Data Retention and Deletion

Your account and project data are retained for as long as your account is active. If you close your account, we will delete your personal information and project data within 30 days, except where retention is required by law. To request deletion, email admin@frontierrnd.com.

Changes to This Policy

We may update this policy as the product evolves. If we make material changes, we will notify you by email or via an in-app notice. The date at the top of this page will always reflect the most recent revision.

Contact

For privacy-related questions or requests, contact Frontier R&D at admin@frontierrnd.com.