AI Policy

Lessona drafts. You decide. Every lesson plan, presentation, and resource is a starting point that you, the qualified teacher, are expected to read, edit, and adapt before using with your class. Lessona does not replace your professional judgement, your knowledge of your students, or curriculum guidance from your school.

We design the product around this assumption. Outputs are always editable. Nothing is locked. The teacher is the final author.

Lessona uses two AI providers: Anthropic for text generation and Google for image generation. Both are major established providers with formal data-handling commitments. We pay them per request. We do not train our own models.

We may change providers or add new ones over time as the field evolves. When we do, we will update this page and email subscribers before any change takes effect.

We have explicitly opted out of model training on every API call we make to our providers. Your lesson topics, classroom context, and generated content are never used to improve a third-party model.

This is enforced at the provider account level and on every individual API request, so the protection applies whether you are on a free trial, a paid subscription, or covered by a school plan.

Lessona is built to plan teaching, not to track learners. We never ask for, collect, or store student names, photos, grades, or any other personally identifying information about your students. The product is structurally incapable of handling student data because we never built the schemas for it.

When you describe your "ability groups" or class context in your profile, those descriptions are about your teaching, not your students. We treat them as pedagogical preferences, not personal data.

We layer multiple safeguards: the providers' built-in safety filters, our own keyword checks on freeform inputs, and structured fields that constrain what can be generated. Blocked requests are logged for compliance review. If something inappropriate ever slips through, tell us via support@lessona.ai and we will address it.

AI can be wrong. It can be out of date, culturally tone deaf, or confidently incorrect about a fact or a curriculum requirement. Treat every Lessona output as a draft. Cross-check anything that must be authoritative (curriculum codes, dates, attributions) with your school resources before using it with students.

Lessona is not a substitute for qualified teaching, professional development, or formal curriculum advice from your ministry or board. We accept no liability for content you choose to use after reviewing it.

For every generation we log who requested it, when, which provider handled it, how long it took, and the estimated cost. We do not log the full text of your prompts as a default. We sample for quality issues only when a generation fails or a teacher reports a problem.

AI is a fast-moving space. If we change providers, models, or how we handle your inputs, we will update this page and email subscribers before any change takes effect.