AI Processing & Transparency Notice

Pre-launch version 0.9 · Last updated 29 August 2026.

1. When AI is used

Bilveo may provide AI-assisted features in paid plans or other clearly labelled features. AI processing should occur only when a user intentionally invokes an AI function or has configured an automated/scheduled feature that clearly discloses its cadence and data use. A deterministic audit is not treated as AI processing merely because Bilveo is an AI-oriented product.

2. Data sent for AI processing

Bilveo follows a data-minimisation approach: only information reasonably necessary to answer the selected analysis task should be sent. Potential inputs can include page URL, relevant metadata/headings, bounded visible text, selected audit findings, user query, domain/brand context and language/location context. Exact production fields and payload limits must be verified before activation.

Full HTML should not be transmitted by default unless a specific feature requires it and the user is adequately informed. Authentication credentials, form secrets, cookies and unrelated browsing data should be excluded from AI payloads.

3. Providers and credentials

The production AI-provider list will appear in the Subprocessor Notice. Bilveo intends to keep provider API keys server-side and prevent the browser/extension from exposing secret provider credentials. Provider retention and model-improvement practices depend on the contracted API/service configuration and must be verified before a provider is enabled.

4. No guaranteed outcome

AI analysis is probabilistic. Scores and recommendations are decision-support outputs. AI Readiness is not evidence that a particular AI service has crawled, indexed, mentioned, cited or recommended a site. Live AI Visibility is a separate measurement concept and should identify provider/product/query/time context.

5. Automated decisions

Bilveo is not intended to make decisions producing legal or similarly significant effects about individuals. A separate impact and legal assessment would be required before such a feature could be activated.

6. Sensitive data

Users should avoid submitting pages containing unnecessary sensitive or special-category personal data. Where technically practical, Bilveo should exclude or redact form fields, authentication areas and obvious sensitive inputs before sending page information to an AI provider.

7. Scheduled monitoring

If scheduled AI Visibility monitoring is introduced, the user must be shown what queries are monitored, which providers are queried, monitoring frequency, credit use and retention of observation records. This feature is not treated as implemented until its backend scheduler and storage are verified.

8. Human review

Users remain responsible for reviewing AI recommendations before making material website, content, legal, security or business decisions.