Last updated: [Effective Date]
Audivance uses AI — powered by Anthropic's Claude — to assist with certain tasks, such as auto-mapping trial balance accounts and suggesting risk or analytical observations. This policy explains how those features work and how they're meant to be used.
AI-generated suggestions are a starting point, not a finished conclusion. Every AI-assisted output in Audivance — account mappings, risk observations, drafted analysis — must be reviewed, edited if necessary, and approved by a qualified member of the engagement team before it is relied upon. Audivance is a tool that supports the auditor; it does not replace the auditor's professional judgment.
When you use an AI-assisted feature (for example, "Auto Map with AI"), the relevant data needed to perform that specific task — such as trial balance line items or financial statement figures — is sent to Anthropic's Claude API to generate a response. We do not send more data than is needed for the feature you triggered.
Data sent to the Claude API for your firm's requests is not used by Anthropic to train their models, consistent with Anthropic's standard commercial API terms.
On plans where AI features are enabled, usage is billed based on the volume of data processed (token usage), on top of your subscription. This is the same billing basis noted directly in the product (see the "Auto Map with AI" tooltip) and on the Pricing page.
AI features are intended for genuine engagement work. Automated, bulk, or scripted use of AI features outside of normal engagement workflows, or use intended to test or probe the underlying AI service rather than to perform audit work, is not permitted.
Questions about how AI features work can be sent to info@getaudivance.com.