1. Purpose
On Your Terms uses artificial intelligence to help you draft legal documents and detect risks. This policy explains how AI is used in our Service, what we do with your prompts and outputs, and the limits of AI-generated content.
2. Where AI is used
- Policy generation — we send your business context to a large-language model to produce draft Privacy Policy, Terms of Service, AI Content Policy, DMCA Policy, Cookie Policy, Acceptable Use Policy, and Data Processing Addendum text.
- Site scanning & drift detection — we use AI to summarize scan findings and rate severity.
- Support & help text — we may use AI to summarize knowledge-base answers.
3. Models and providers
By default we route requests through the Lovable AI Gateway to Google Gemini models (currently google/gemini-3-flash-preview). The default model may change with notice. Where another provider would be used, we will disclose it.
4. What we send to the model
Only the business context you provide for the relevant action (company name, website URL, contact email, industry, jurisdiction, the existing policy text being revised, the page HTML being analyzed). We do not send your full account history.
5. Training
We do not use your prompts, generated outputs, or your customers' data to train base models. Our AI provider's terms apply to data sent to the gateway and currently prohibit use of API content for training. We will update this section if those terms change.
6. Retention
Prompts and outputs are stored with your account so you can view, version, edit and approve drafts. You can delete them at any time. We may retain anonymized logs of model latency and error rates for operational monitoring.
7. Accuracy and limits
AI output can be inaccurate, incomplete, out of date, or biased. Legal language depends on jurisdiction and facts the model cannot verify. Treat every generated document as a starting draft that requires:
- review against your actual business practices;
- review by qualified counsel licensed in the relevant jurisdiction(s);
- adaptation to your specific risk profile.
We disclose this prominently in the product and on every generated document.
8. Prohibited uses
You will not use our AI features to:
- generate content designed to deceive consumers about their legal rights;
- produce illegal content (CSAM, incitement, defamation, fraud);
- impersonate a real person or organization;
- infringe intellectual property;
- circumvent safety filters or extract another party's proprietary information.
9. Reporting concerns
Report harmful, infringing, or unsafe model output to trust@onyourterms.app. We investigate within 5 business days.
10. Changes
Material changes to this policy will be notified by email and an in-product notice.
Important Notice
AI-generated output is not a substitute for legal advice. Always have qualified counsel review documents before publishing.