Assenture

AI usage policy

Last updated: June 17, 2026

Assenture includes AI assisted features ("AI Features") that help you classify transactions, summarise documents, draft communications, and answer questions about your workspace. This policy explains how those features work, what data they process, and the limits we place on them. It applies in addition to our Terms of Service, Privacy Policy, and Data Processing Addendum.

1. Models we use

AI Features are powered by third party foundation models accessed through the Lovable AI Gateway, currently including Google Gemini and Anthropic Claude families. The specific model is selected by us based on quality, latency, and cost, and may change. The current active providers are listed on the Subprocessors page.

2. What data is sent to the model

When you use an AI Feature, we send the minimum data necessary for the request, which may include:

  • Your prompt or instruction.
  • Relevant rows from your workspace (for example, transactions you ask the assistant about).
  • The contents of files you explicitly attach to a chat.
  • A short conversational history to maintain context within the same session.

We do not send your password, payment card details, or other customers' data. We do not send your data to AI providers for any purpose other than answering your request.

3. No training on your data

We have contractual commitments from our AI subprocessors that customer data submitted to their APIs is not used to train or improve their foundation models, and is not retained beyond the period required to deliver the response and meet abuse-monitoring obligations (typically 30 days or less, depending on the provider).

4. Human oversight and "always allow" actions

The AI assistant can navigate the application and propose actions on your behalf (creating drafts, classifying items, generating reports). By default these actions require your explicit confirmation. You may mark specific action types as "always allow" for your own user account; that consent is recorded, scoped to you, and can be revoked at any time from your settings. We never apply "always allow" across users.

5. Limitations and disclaimers

  • AI Features can produce inaccurate, incomplete, or misleading output ("hallucinations"). You are responsible for reviewing AI output before relying on it, especially for financial reporting, tax filings, or legal decisions.
  • AI output is not professional advice. See our Financial Disclaimer.
  • We do not guarantee that the same prompt will always produce the same answer.

6. Prohibited uses

You may not use AI Features to:

  • Generate content that is unlawful, defamatory, discriminatory, harassing, or sexually exploitative of minors.
  • Attempt to extract training data, system prompts, or other users' data from the model.
  • Make consequential decisions about individuals (employment, credit, insurance, housing) without meaningful human review.
  • Generate disinformation, impersonate real people, or evade detection systems.

7. Logging

We log AI prompts and responses tied to your workspace for up to 90 days for debugging, abuse prevention, and to let you review what the assistant did. Admins of your workspace can read these logs; we cannot, except where strictly necessary to investigate a reported issue or to comply with law.

8. Turning AI off

Workspace admins can disable AI Features entirely from Settings → AI. When disabled, no data is sent to any AI provider from that workspace.