Privacy and your data

This page explains what information Rootpublish's free beta stores, why it stores it, and how it can be shared.

Operator

Nandarona Inc. (株式会社なんだろな) / Rootpublish

Questions about the service or personal information: support@rootpublish.com

Why we use information

We use information to verify your identity and maintain sign-in; save, collaborate on, draft, publish and export content in workspaces; answer inquiries; prevent misuse; and recover from incidents. Aggregate usage counts help us improve confusing interactions and features. We do not use your information to serve advertising or sell the content you enter.

Information we store

We store your display name, email address and account identifier, workspace membership, conversations, sources and drafts you enter, and their sharing settings, provenance and change history. Google verifies your identity and provides your email address, basic profile and account identifier. We do not request access to Gmail message bodies, Google Drive files or search history. Sign-in uses Better Auth, with authentication records and sessions stored in Cloudflare's database. Necessary cookies maintain sign-in, and sessions expire. Google connection tokens are stored encrypted. Sessions may include an IP address and browser identification information. The local test version uses a database on the current device.

Storage and security

The site and database use services provided by Cloudflare, a US company. The primary database is in the APAC region and is currently located in Singapore. Communications and service processing also use facilities outside Japan. We use encrypted communications, account and workspace access controls, encryption for authentication tokens and API keys, request limits, and settings that exclude private content from ordinary application logs. Please also read Cloudflare's policy and Google's policy for their handling of information.

Sharing and publication

New sources and original conversations are visible only to their author. You can separately choose workspace sharing and permission to use a source in external content. You choose whether to display a draft on a public page after approving it. Search engines and third parties may retrieve and keep public content. Unpublishing does not guarantee deletion of copies already obtained.

Sending information to AI

The “Assemble original words” method does not send content to an external AI. If you choose “Write with your AI,” review what will be sent, and press the dedicated send button, the titles and bodies of selected sources that are reviewed, shared and cleared for publishing, together with the output format, are sent to your configured OpenAI, Anthropic or Google Gemini provider. Temporary numbers link the sources; account information, original conversations and unselected sources are excluded. API keys are encrypted and can be deleted by the owner. Workspace owners and editors can make requests with the key. Retention, use of information after sending, and AI charges follow your provider's agreement and settings. Disconnecting cannot retract information already sent. For AI connections in the standalone WordPress plugin, consult its settings and explanations.

Usage measurement

On public marketing pages, we load Google Analytics 4 through Google Tag Manager only after you allow analytics. Cookies measure page views and clicks to start, download, switch language or contact us. Event fields are limited to a public page URL, fixed page title, interface language, page type, action and button placement. Query strings, URL fragments, referrer URLs, email addresses, invitation tokens and workspace content are excluded. Google measurement tags are not loaded on sign-in, workspace or public profile pages. Google also processes technical device and browser information. We do not use advertising personalization or Google signals.

You can use the service without consenting. Change your choice with “Cookie settings” at the bottom of the page. The choice is stored in this browser for 180 days. Withdrawal stops new Google measurement and deletes this site's GA cookies; it does not automatically delete previously submitted data. See Google's privacy policy for Google's handling of information.

We record landing-page views and clicks on start and download buttons as aggregate counts by date, event and placement. Events do not contain conversation text or email addresses. Temporary hashes help limit requests to prevent excessive access. We do not use advertising cookies for measurement.

AI facts check connection (MCP)

When an assistant such as ChatGPT connects to https://rootpublish.com/mcp for the AI facts check, the site address, the assistant's answers and its judgements are used only for that check and are not stored. To compare them, we read the site's public pages and the public pages the answers cite, at most once a second and following robots.txt, and only from public https addresses. Temporary hashes of IP addresses help limit requests.

Export and deletion

Workspace settings let you export your conversations and the sources and drafts you can access. Archiving a source is not permanent deletion. You can delete your Rootpublish data from your personal workspace settings. First remove other members from any workspaces you own. Your Rootpublish login and sessions on all devices are deleted at the same time. This does not delete your Google account, public information obtained by third parties, or text already quoted in another person's draft. Contact the operator above for deletion inquiries. Requests for access, correction, deletion or suspension of use are handled after identity verification in accordance with applicable law.

Retention and backups

Accounts and created content are retained until you delete them. Archiving is not permanent deletion; change history and records needed for collaboration are also retained. Login sessions last seven days. Expired sessions and temporary authentication records are cleaned up daily. Aggregate usage and operation audit records older than 90 days are deleted daily.

When account deletion completes, the relevant data is removed from the active database. Cloudflare disaster-recovery backups may retain information from before deletion for up to 30 days. If the contracted recovery window is seven days, that shorter window applies. Backups are not used for ordinary page display or sales purposes.

To prevent deleted data from returning through an old backup, we retain internal identifiers and deletion timestamps without content or email addresses. Records older than 35 days are deleted daily. These records themselves may remain in backups for up to another 30 days. During restoration, we stop the service and reapply deletion records. We revoke previous publication and sharing, sessions and AI connections before resuming. Temporary recovery copies are deleted after verification, with a target of clearing them within seven days after incident response ends.

Changes and contact

We announce material changes to data handling on this page or elsewhere in the service. Contact the operator above with questions.

Policy updated: September 14, 2026