Privacy Policy

AI Reserve (operated by Audacity Advisory Corp) provides an enterprise AI platform: organizations fund a wallet and their people use AI models from many providers through one gateway, one set of keys, and one ledger. This policy explains what we collect, why, and the controls you and your organization have. Where your organization has signed a Terms of Service or Data Processing Addendum with us, those agreements govern.

• Account information — name, work email, organization, role, and authentication identifiers.
• Usage metadata — for every request: model, token counts, cost, latency, status, and which user, key, and team it belongs to. This is billing and spend-governance data; it contains no message text.
• Prompt and response content — by default we store message text to power chat history, search, audit, and usage analytics. Your organization can turn this off entirely (see "Content storage controls").
• Documents — files you or your organization upload for use with AI features.

• Providing the service: routing requests, chat history, document features.
• Metering and billing: every request is attributed and counted against your organization's budgets.
• Security and abuse prevention: authentication, rate limiting, anomaly detection.
• Usage analytics for your organization's administrators, including optional business-function classification of usage. Classification reads message text to categorize spend; enterprise administrators can disable it, and it is automatically off when content storage is off.

We do not use identifiable customer data to train AI models, and we do not sell personal information.

Enterprise administrators can disable prompt and response storage for their whole organization ("Store prompt & response content" in the admin console). While off, no message text is persisted — only the usage metadata above — and content-based classification stops. Administrators can also permanently delete all previously stored content for their organization on demand.

Your prompts are processed by the AI provider serving the model you chose (for example OpenAI, Anthropic, Google, or AWS Bedrock). Each provider's retention and training posture differs; we document them per provider on our Data Handling page, and we list all subprocessors on the Subprocessors page.

Data is encrypted in transit and at rest. Access is role-based and scoped to your organization (tenant isolation). The platform runs on infrastructure certified to SOC 2 Type II and ISO 27001 (Google Cloud). Requests are processed in the United States.

• Delete your data — from your profile you can delete all your chats and documents while keeping your account, or delete your account entirely.
• Organization-level deletion — enterprise administrators can purge all stored message content for their organization.
• Access and correction — contact your organization's administrator or our team to access or correct your account information.

Depending on where you live, you may have additional rights under laws such as GDPR or CCPA/CPRA; we honor verified requests consistent with those laws.

Usage metadata is retained for the life of the account for billing integrity and audit. Stored message content remains until you or your organization deletes it, or immediately stops accruing when content storage is turned off. Deleted content is removed from production systems promptly and from backups on their rotation schedule.

We will update this page when our practices change, and material changes will be communicated to organization administrators. Questions or requests: contact your AI Reserve account team through the portal.