Tivoraly
Log In
Policy

AI Transparency

Where AI is used, what it can do, and what it does not guarantee.

Version 1.2Last reviewed 2026-09-18Replaces version 1.1 Applies globally

This wording is pending professional legal review. It describes how the platform operates and is not legal advice.

Preview wording revision · 2026-10-29. The stored policy version, effective date and acceptance records remain unchanged. This revision is not yet published on the live website.

In plain English: Some Tivoraly features use AI to draft or assist. Outputs can be wrong, incomplete or unsuitable; a parent must review them. AI is not a complete curriculum or legal/safety guarantee.

Current uses

AI-assisted paths include requested Pro learning preparation/planning, activity import and refinements, saved-activity step assistance and optional report summaries. Other paths include Help assistance, course-description assistance, nearby mission drafting, internal research and group-listing moderation where enabled. Availability and access limits depend on the feature.

Manual activities and Planner entries, worksheet references, rule-based practice worksheet generation, PDF annotation and ordinary Arcade gameplay do not themselves require AI generation. Retired paste-to-worksheet and project/lesson generator entry points are not current tools.

Information sent

Requests can include your prompt, relevant saved learner nickname/age/year, interests, engagement/learning preferences, dislikes/support notes, activity content, conversation context, selected report records, or text/images/files you submit for processing. Relevant saved context can be assembled by the application, not just typed by you. Not every feature uses every category.

Requests are routed through the Lovable AI Gateway to model providers. Some offered saved-activity voice input may use an external transcription fallback. Device speech uses browser/device services where offered; do not assume all speech is offline.

Unresolved provider facts: final processing regions, retention, training use, human access and deletion terms need confirmation. No unverified no-training/no-retention promise is made. Avoid unnecessary identifying or sensitive material. See Privacy Policy.

Parent review and reporting

Review AI material for accuracy, suitability, omissions, supervision and safety. The tools are instructed to preserve safety guidance, but automated instructions and labels are not guarantees. Source/AI labels are shown where a feature provides them; not every newly discovered result has been manually reviewed.

Listing screening is content moderation, not organiser identity, qualifications, registration or child-safety verification. Use available Report controls or Contact Us to raise problems. Google Play's AI-content and child-audience requirements still need a release-specific assessment.

Loading…