AI transparency

Last updated: 2026-10-20

What AI does

Didavero uses generative systems to help research sources, structure and write notes, draft exams and propose teaching programmes, units and activities. It assists teaching-material preparation: results are AI-generated proposals, not official rules or decisions about a learner. The teacher decides what to keep, edit, activate or export.

Engine and provider

The app integrates user-configured tools: Claude Code, Codex or OpenCode. The effective model depends on settings and account; no single model is represented as covering every installation. Jobs may consume quotas or tokens billed by third parties. This informational website has no AI chatbot and does not request visitor documents.

Limits and oversight

AI can fabricate references, make calculation errors, omit content, introduce bias or misinterpret dates and regional rules. Check official sources, curriculum references, hours and outputs before using materials. A citation or automated check does not prove accuracy, completeness or administrative compliance. Keep sources and record your review decisions.

Data and educational use

Use fictional examples or anonymised material. Do not enter identifiable children’s data, health, disability or other sensitive categories without an appropriate legal basis, safeguards and institutional authorisation. The intended use does not include grading real learners, admissions, educational placement or exam surveillance. Repurposing it for those purposes requires a separate legal and technical assessment; human involvement alone does not remove a possible high-risk classification.

Generated content

Identify AI involvement when sharing or publishing outputs and preserve available provenance. Do not remove origin signals or metadata provided by the model provider. The app does not guarantee that every export contains machine-detectable AI marking. Applicable transparency and marking duties must be checked for the distribution, provider and specific use; this page does not replace those technical measures.

Literacy and incidents

Before using AI, understand its limits, provider settings, data protection and output review. Institutions should organise the literacy and oversight measures relevant to them. Report significant errors, bias or risks to pablo@didavero.com with version and reproduction steps, using fictional examples without learner data. No AI Act conformity certification is advertised.