Legal

RefNode AI Privacy Policy

Last updated: October 29, 2026

This Privacy Policy explains how this site, operated by Hyperscale Dynamics Corporation dba Reference Node (“RefNode,” “we,” “us,” or “our”), collects, uses, discloses, retains, and protects information when you use RefNode AI.

By using RefNode AI, you acknowledge this Privacy Policy. If you do not agree, do not use the Service.

1. Information We Collect

We may collect the following categories of information.

Account Information

This may include your name, email address, username, password or authentication credentials, account settings, subscription status, and related account details.

Payment Information

If you purchase a subscription, credits, or paid features, payment information may be processed by a third-party payment provider. We may receive limited billing details such as payment status, transaction identifiers, plan type, and invoice history. We do not intend to store full payment card numbers.

User Content

We may collect prompts, questions, files, uploads, instructions, context, chat history, council configurations, model selections, outputs, feedback, and other content you submit to the Service.

Saved Conversation History (RefNode Pro and Max)

This cloud-history feature does not apply to Investment Research. Its research history remains local to your browser on the current device, regardless of subscription tier.

If you subscribe to RefNode Pro or Max, your council conversations are saved automatically so you can revisit them and continue them across your devices. Saved conversations, including their titles, are encrypted at rest (AES-256-GCM) before being written to storage — the storage provider holds only ciphertext. Saved conversations are linked to your account and are not accessible to any other user. Our staff does not access saved conversations in the ordinary course of operating the Service.

You are in control of this history: you can delete any saved conversation, or all of them, at any time from the History sidebar, and deletion removes the stored data. History keeps your 100 most recent conversations (older ones are automatically pruned), and very long conversations may have their earliest exchanges trimmed from the saved copy to fit storage limits. Note that conversation content is always processed by the AI model providers that generate the answers (see Section 3), whether or not history is stored.

Usage Information

We may collect information about how you use the Service, including timestamps, feature usage, council usage, model-routing events, latency, errors, logs, token usage, approximate usage volume, account activity, and diagnostic information.

Device and Technical Information

We may collect IP address, browser type, device type, operating system, referring pages, session identifiers, cookies, and similar technical information.

Previously Connected Credentials

RefNode AI no longer accepts personal API keys or other provider credentials. If you connected a key under a previous version of the Service, RefNode AI does not read or use it. A copy may remain in your own browser’s storage until you clear site data for refnode.us. On request to help@refnode.us we will delete any encrypted copy stored with your RefNode account. We recommend revoking any such key in your provider account.

Communications

If you contact us, we may collect your message, email address, contact information, support request, and any information you choose to provide.

2. How We Use Information

We may use information to:

3. AI Processing, Model Routing, and Third-Party Providers

RefNode AI is independently operated by Hyperscale Dynamics Corporation dba Reference Node. It provides access to models from OpenAI, Anthropic, Google, and other AI developers; it is not operated by those model developers. We use OpenRouter to route AI requests to model providers through encrypted HTTPS API connections. Learn more in OpenRouter’s privacy policy.

RefNode AI may route prompts, context, files, outputs, metadata, usage information, and related request information to third-party model providers, infrastructure providers, model routers, authentication providers, payment providers, analytics providers, and other service providers.

These providers may process information according to their own terms, privacy policies, retention practices, logging practices, security controls, and model policies.

4. Training and Product Improvement

Unless otherwise stated in a separate written agreement or account setting, RefNode AI may use usage data, diagnostics, feedback, and interaction metadata to maintain, secure, debug, and improve the Service.

RefNode AI should not use your private prompts, uploaded files, or private chat content to train public foundation models unless we expressly disclose that practice or obtain appropriate permission.

Third-party model providers may have separate practices regarding logging, retention, abuse monitoring, and model improvement. RefNode AI selects and contracts with these providers; their policies apply to the requests we route to them on your behalf.

5. Sensitive and Prohibited Information

RefNode AI is not designed for classified information, Controlled Unclassified Information, export-controlled technical data, ITAR-controlled data, EAR-controlled technical data, government-sensitive information, procurement-sensitive information, protected health information, financial account numbers, Social Security numbers, passwords, private keys, trade secrets, or other restricted information.

Do not submit prohibited or sensitive information unless RefNode AI has expressly agreed in writing to support that use case.

6. How We Disclose Information

We may disclose information to:

7. Cookies and Similar Technologies

For conversion measurement, we use a random browser-session identifier lasting up to 24 hours and broad device, browser, and referral categories. This helps us count visits, submitted questions, completed answers, plan views, checkout starts, and confirmed payments. These measurements exclude question and answer text, email addresses, full URLs, and raw referrers. The identifier and categories may accompany a checkout in Paddle so its payment confirmation can be counted. Our conversion counters and deduplication records expire after 15 days; Paddle retains transaction metadata under its own policies.

We may use cookies, local storage, pixels, or similar technologies to operate the Service, keep you logged in, remember preferences, measure usage, detect abuse, and improve the Service.

You may be able to control cookies through your browser settings. Some features may not work properly if cookies are disabled.

8. Data Retention

We retain information for as long as reasonably necessary to provide the Service, maintain accounts, comply with legal obligations, resolve disputes, enforce agreements, detect abuse, maintain security, and improve the Service.

Saved conversation history (RefNode Pro and Max) remains stored until you delete it from the History sidebar. Automatic pruning applies past the 100 most recent conversations.

Retention periods may vary based on the type of information, account settings, legal requirements, provider requirements, operational needs, and security needs.

Third-party providers may retain information according to their own policies.

9. Security

RefNode AI uses HTTPS to encrypt connections between your browser and our service, and between our service and the AI routing API. The services processing a request receive its content to generate an answer; encryption in transit does not prevent that processing.

Our hosting provider, Vercel, and sign-in provider, Clerk, maintain SOC 2 Type II reports covering their services. These independent assessments evaluate controls over a period of time. They apply to the providers’ audited services and do not constitute a separate SOC 2 assessment of RefNode AI or every model provider.

AI providers’ retention, logging, and data-use practices vary, as described in Sections 3 and 4. Our use of encrypted connections and providers with SOC 2 reports does not imply that all request data is immediately deleted or excluded from every provider’s model-improvement practices.

We use reasonable administrative, technical, and organizational measures designed to protect information. However, no system is perfectly secure.

You are responsible for securing your account, protecting your credentials, and avoiding submission of sensitive or prohibited information.

10. International Processing

Information may be processed in the United States and other locations where RefNode AI, its providers, or third-party services operate. Data-protection laws may differ from those in your jurisdiction.

By using the Service, you understand that information may be processed outside your state, province, or country.

11. Your Choices

Depending on your account settings and applicable law, you may be able to:

To make a request, contact us at help@refnode.us.

12. California Privacy Notice

California residents may have rights under California privacy law, depending on how the law applies to RefNode AI. These rights may include the right to know, access, correct, delete, and limit certain uses or disclosures of personal information, and the right not to be discriminated against for exercising privacy rights.

We do not intend to sell personal information. If we engage in activities considered a “sale” or “sharing” under applicable California law, we will provide required notices and choices.

To exercise applicable California privacy rights, contact us at help@refnode.us.

13. Children

RefNode AI is not intended for children under 18. We do not knowingly collect personal information from children under 13. If you believe a child has provided personal information to us, contact us at help@refnode.us.

14. Enterprise, Government, and Regulated Use

Unless we expressly agree in writing, RefNode AI is not intended for use with regulated, classified, controlled, export-controlled, government-sensitive, health, financial, or other restricted information.

Organizations are responsible for determining whether their use of RefNode AI complies with their legal, contractual, procurement, security, confidentiality, retention, residency, and regulatory obligations.

15. Changes to This Privacy Policy

We may update this Privacy Policy from time to time. The “Last updated” date indicates when the Privacy Policy was last revised.

If we make material changes, we may provide notice through the Service, by email, or by other reasonable means.

16. Contact

For privacy questions or requests, contact:

Hyperscale Dynamics Corporation dba Reference Node
[MAILING ADDRESS OR REGISTERED AGENT ADDRESS]
help@refnode.us

Investment Research

RefNode does not maintain a server-side archive of your Investment Research conversations. Research questions, responses, and research records are stored locally in your browser on the current device, where you can delete or download them. They are not cloud-synchronized, including for RefNode Max subscribers.

RefNode may retain limited account, disclosure-acceptance, security, abuse-prevention, and operational records necessary to operate and protect the service. These records do not include an Investment Research conversation archive. The current research service stores your acknowledgment and a daily source-preparation counter separately from conversation content.

Research questions and source text are sent to AI providers through OpenRouter using RefNode's provider account and your plan's AI credits. Automatic searches send the question to the search service through OpenRouter. Optional deeper-review preparation sends your question and public source links to RefNode for transient processing and source retrieval; the resulting preparation is returned to your browser without creating a server-side conversation archive. No personal API key is required or included in research records. OpenRouter and other model, search, and infrastructure providers process requests under their own logging, retention, and privacy practices. Local storage does not mean that no third-party records exist.

Do not enter brokerage passwords, account credentials, Social Security numbers, financial account numbers, or other sensitive financial information.

See the Investment Research Disclosures for product boundaries and limitations.