Worthic uses AI to assist with document processing, categorization, summaries, insights and workflow support. Because financial and administrative records can be sensitive, AI features are designed around privacy, evidence and human review.
AI assists—it does not decide for you
AI-generated results may be incomplete, inaccurate or unsuitable for your circumstances. Worthic therefore treats AI output as assistance rather than final authority.
Where AI affects documents, transactions, categories, routing or reports, Worthic aims to keep the result reviewable. Depending on the workflow, you can inspect, correct, approve, reject or override a suggestion. Actions such as creating pending transactions from an uploaded document may require explicit confirmation.
AI output is not accounting, tax, audit, legal, investment, financial or compliance advice. Verify important figures and classifications, particularly where they affect tax, legal obligations, financial statements, large transactions, investments or business decisions.
Your data is not used for general AI training
Worthic does not use customer documents, financial data or workspace content to train third-party foundation models or general AI models unless the customer explicitly opts in. Opt-in is not automatic and requires separate, explicit consent.
Worthic may use aggregated or de-identified operational information to monitor reliability, performance, cost and error rates, provided it does not identify a person or customer workspace.
What may be processed
When an AI feature is used, the relevant workflow may process:
- uploaded financial or administrative documents;
- extracted text and document metadata;
- transaction descriptions, amounts and related context;
- workspace information needed to answer a question or make a suggestion;
- prior corrections, confirmations or rejections; and
- results produced by rules-based processing.
Worthic uses approved service providers where necessary to deliver AI processing. Data use remains limited to operating the relevant feature and is subject to Worthic’s privacy, security and data-processing controls.
Permissions and workspace separation
AI features follow the same workspace roles and access restrictions as the rest of Worthic. The AI assistant can only retrieve information that the current user is permitted to see. It cannot use chat to bypass access to restricted accounts, reporting lines, documents or administrative functions.
Production, staging and development environments are kept separate. Production processing uses production-specific storage, queues, credentials and service endpoints to reduce the risk of customer information being mixed with test systems.
Human access and auditability
Worthic does not routinely access customer documents for support. Limited authorized access may occur where necessary for troubleshooting, security, abuse prevention, compliance or a support request. Such access is subject to internal controls and auditable administrative-access logs.
If you voluntarily send a document, screenshot or export to support, Worthic may process that material to resolve your request.
Worthic also keeps operational and security logs for areas such as authentication, permissions, document-processing state, errors and administrative access.
Retention and deletion
If a workspace is deleted, Worthic retains its data and documents for 14 days to allow export and recovery. After that, production data is scheduled for deletion or de-identification, subject to limited retention in backups, logs, audit records and information required for security, legal or compliance purposes.
What you can do
You can help protect your workspace by using strong credentials, securing your devices and email account, assigning permissions carefully, removing users who no longer need access and reviewing AI-assisted results before relying on them.
For further detail, refer to Worthic’s Privacy Policy, Terms of Service, Security Overview and AI and Data Processing Statement.