AI Transparency

TeachAIFlow is a generative AI platform built for teachers. It produces drafts for exams, presentations, lesson plans, rubrics, family messages and other classroom resources based on the instructions you provide.

Every AI output is a draft. The final decision, evaluation, correction and communication to students or families always remains with the teacher. TeachAIFlow does not replace the teacher's professional judgment and does not issue official grades on its own.

Yes. In line with Article 50 of Regulation (EU) 2024/1689, TeachAIFlow tells you explicitly that you are interacting with an AI system. This indication appears within each tool and as a watermark on PDFs generated under the free plan.

AI-generated content (text, questions, slides, rubrics, etc.) may contain errors, inaccuracies or biases. You must review it before using it in class.

Under the EU AI Act, TeachAIFlow is classified as a limited-risk AI system in its default uses. This means the applicable obligations are fundamentally transparency obligations toward the end user (this page), and the system does not require CE conformity or registration in the European high-risk AI database.

There are concrete scenarios where a specific use could approach a high-risk system under Annex III of the Regulation (e.g., systems used to officially evaluate students or decide their admission). TeachAIFlow is not offered as an official evaluation tool: the exam-correction tool produces an editable draft that the teacher reviews, modifies and signs as their own.

TeachAIFlow uses generative AI models from external providers (specifically OpenAI). We do not train or store our own models.

Specific models may change to improve quality, latency or cost. Current providers and models in use:

• OpenAI — GPT models (gpt-4.1-mini and gpt-4.1-nano families by default). They process your instructions and the texts you upload to generate the response. OpenAI's enterprise API policy is not to train its models on data submitted through the API.
• Pexels and Unsplash (stock-image providers) — used by the presentation generator when illustrative images are chosen. They do not process your personal data.

When you use a tool, the following data may be sent to our AI provider (OpenAI) to generate the response:

• The text you type as instruction or topic.
• The documents (PDF, DOCX) you upload to the tool, as extracted text.
• Non-identifying system metadata (language, number of questions requested, target length).

We protect your privacy as much as possible. In particular, we DO NOT send to OpenAI or any other AI provider:

• Your name, email or any personal identifier.
• Personal data of your students (real names, grades, sensitive data).
• Payment information (Stripe processes payments directly; AI models have no access).
• Your current plan, credit usage or financial history.

TeachAIFlow is designed under the principle of meaningful human oversight (Article 14 of the Regulation). In practice:

• Every generated draft is editable before exporting to PDF or Word.
• Exam corrections are drafts that the teacher must review and validate before communicating to the student.
• Lesson plans, rubrics and family communications are proposals, not official documents issued by the platform.
• We never send content directly to your students or families without your intervention.

To make informed decisions, you should understand the limitations of the generative AI systems powering TeachAIFlow:

• Hallucinations: AI may invent facts, dates, quotes or bibliographic references. Always verify before including in academic material.
• Bias: models inherit biases from their training data. Review content for neutrality and representation.
• Limited knowledge: models have a knowledge cutoff. Information after that date may be inaccurate.
• Language: performance can vary across languages; Spanish and Catalan quality is optimized but not infallible.
• Local curricula: drafts are calibrated to the detected country curriculum (Spain, Mexico, Argentina, etc.), but should be reviewed for exact alignment with your school's syllabus.

In line with the spirit of Article 50.2 of the AI Act, PDFs and materials generated under the free plan include a visible watermark reading “TeachAIFlow FREE”. PRO users can export without the watermark, but the material is still AI-generated and must be identified as such if it will be published, distributed to families, or used officially.

The teacher remains responsible for declaring the use of AI when required by their school's or education authority's rules.

The exams, presentations and materials you generate are stored in your account so you can reuse them. You can delete them at any time from your workspace.

Intermediate instructions sent to the AI provider are not stored on our servers after generation; the result is.

As a user, you have the right to:

• Be clearly informed of when you interact with AI (this page and the in-product notices).
• Know the system's limitations (section 8).
• Request an explanation of how a specific tool works.
• Request deletion of generated content stored in your account.
• Not have your students subjected to a decision based exclusively on an automated system (Art. 22 GDPR); TeachAIFlow structurally guarantees this by requiring teacher review.

If you encounter an AI output that you consider incorrect, harmful, biased or unsafe, report it to teachaiflowapp@gmail.com. We treat every incident as input to improve the tools and, where appropriate, adjust the system prompts or the documentation.

TeachAIFlow does not share specific user outputs for product improvement without prior consent. Quality analyses rely on aggregated, anonymized metrics.

We will update this transparency policy when we add new models, expand the tools or applicable legal obligations change. The date shown in the header reflects the current version.

For questions about our AI use, write to teachaiflowapp@gmail.com. If you believe your rights as an AI user or your data-protection rights have been violated, you can contact the Spanish Data Protection Agency (AEPD) or the national authority designated to supervise the AI Act in your country.