AI Usage Disclosure

Tutor Volt Ltd | Last updated: 15 February 2026

1. Introduction

Tutor Volt uses artificial intelligence (AI) as a core part of its educational platform. We believe in being transparent about how AI is used, what its limitations are, and what choices you have. This document explains everything you need to know about AI on Tutor Volt.

2. What AI Is Used For

The following table summarises every feature on the Platform that uses AI:

| Feature | What It Does | AI Role | |---|---|---| | Lesson Content Generation | Creates educational lesson content aligned to the UK national curriculum | AI generates initial lesson text, diagrams, and examples. A qualified teacher reviews and approves every lesson before it is published to students. | | AI Tutor Chat (Live Wire) | Provides real-time, interactive learning support to students | AI answers student questions, explains concepts, provides worked examples, and offers encouragement. All conversations are monitored for safeguarding. | | Quiz Generation | Creates quizzes and assessments based on lesson content | AI generates questions, answer options, and mark schemes. Teachers can review and edit before assigning to students. | | Flashcard Generation | Creates revision flashcards from lesson content | AI extracts key terms, definitions, and concepts to create spaced-repetition flashcards. | | Mind Map Creation | Generates interactive mind maps showing concept relationships | AI analyses lesson content and creates structured mind maps showing how topics connect. | | Grade Predictions | Provides predicted grades based on student performance data | AI analyses historical quiz scores, completion rates, and learning patterns to generate indicative grade predictions. These are advisory only. | | Content Quality Scoring | Assesses the quality and accuracy of AI-generated content | AI evaluates generated content against curriculum standards and flags potential issues for teacher review. | | Safeguarding Detection | Monitors student interactions for safeguarding concerns | AI analyses all student chat messages in real-time using keyword patterns and contextual understanding to detect potential safeguarding concerns. See our Safeguarding Policy for details. | | Adaptive Learning | Adjusts content difficulty based on student performance | AI tracks student performance and adjusts quiz difficulty, content recommendations, and learning paths accordingly. |

3. The AI Model

3.1 Which AI

Tutor Volt uses Claude, an AI model developed by Anthropic (San Francisco, United States). Claude is a large language model designed with a focus on safety and helpfulness.

3.2 Additional AI Services

• Deepgram (United States) -- Provides speech-to-text processing, converting spoken words into text for the voice interaction feature.
• ElevenLabs (United States) -- Provides text-to-speech audio generation, converting lesson text into natural-sounding spoken audio.

3.3 How AI Processes Data

When you interact with AI features on the Platform:

• Your question or input is sent to the AI service for processing
• The AI generates a response based on the input and its training data
• The response is returned to the Platform and displayed to you

Personal identifiers (your name, email, school) are not sent to the AI model. We send only the educational content needed to generate a useful response (e.g. your question, the lesson topic, the curriculum context).

4. Human Oversight

AI on Tutor Volt is designed to support human teaching, not replace it. We maintain human oversight at every stage:

4.1 Content Review

• All AI-generated lesson content is reviewed by a qualified teacher before it is published to students. No student sees AI-generated lesson material that has not been approved by a teacher.
• Teachers can edit, modify, or reject any AI-generated content.
• AI-generated quizzes and flashcards can be reviewed and edited by teachers before assignment.

4.2 Safeguarding Review

• Automated safeguarding alerts are reviewed by our Designated Safeguarding Lead (DSL), who is a trained human professional.
• No safeguarding action is taken based solely on an automated decision. All critical and high-severity concerns are reviewed by a human before any intervention.

4.3 Grade Predictions

• AI-generated grade predictions are presented as advisory indicators only.
• Teachers may use predictions as one factor among many in their professional judgement.
• No educational decision with legal or similarly significant effect is made solely on the basis of AI-generated predictions.

5. What Data Is Sent to AI

We are transparent about what data the AI receives:

| Data Sent | Purpose | Identifiable? | |---|---|---| | Student questions (AI chat) | To generate helpful educational responses | No -- questions are sent without the student's name or personal identifiers | | Lesson topics and curriculum specifications | To generate curriculum-aligned content | No -- contains only subject matter, not personal data | | Quiz answers (for adaptive learning) | To adjust difficulty and recommend content | No -- answers are processed in context of the lesson, not linked to personal identity in the AI request | | Voice recordings (speech-to-text) | To convert spoken input to text | Minimised -- audio is processed and deleted within 30 days. No name or identifier is sent with the audio. | | Chat messages (safeguarding) | To detect safeguarding concerns | The message content is analysed. No personal identifiers are sent to the AI for safeguarding analysis. |

6. AI Limitations

We believe it is important to be honest about what AI can and cannot do:

6.1 AI May Generate Inaccurate Content

AI models, including Claude, can produce information that is incorrect, incomplete, or misleading. This is sometimes called a "hallucination." While we take steps to minimise this (including teacher review), no AI system is perfectly accurate.

6.2 AI Is Not a Substitute for Human Teaching

The AI tutor (Live Wire) provides educational support, but it is not equivalent to instruction from a qualified teacher. It cannot assess a student's emotional state, adapt to classroom dynamics, or provide the full range of support that a human teacher offers.

6.3 AI Has Knowledge Limitations

The AI model's knowledge has a training cut-off date and may not reflect the very latest curriculum changes, research findings, or current events.

6.4 AI Cannot Make Professional Judgements

AI-generated grade predictions, content quality scores, and learning recommendations are statistical assessments, not professional educational judgements. They should always be interpreted by qualified educators.

7. Your Choices

You have control over how AI is used in your or your child's learning:

7.1 AI Tutor Chat (Live Wire)

• Parents can disable the AI chat feature for their child through the parent dashboard settings.
• When disabled, students will still have access to pre-published lesson content, quizzes, and flashcards, but will not be able to have interactive conversations with the AI tutor.

7.2 Grade Predictions

• Parents and students (aged 13+) can opt out of AI grade predictions at any time.
• To opt out, contact [email protected] or use the privacy settings in the Platform.
• Opting out means the Platform will not generate or display predicted grades for that student. It does not affect access to other features.

7.3 Voice Features

• Speech-to-text and text-to-speech features are optional. Students can choose to type instead of speak, and can read lesson content instead of listening to audio.
• Parents can disable voice features through the parent dashboard settings.

7.4 Adaptive Learning

• The adaptive learning system adjusts content difficulty based on performance. This is integral to the Platform's educational approach.
• If you have concerns about how adaptive learning affects your child's experience, contact us at [email protected].

8. No Automated Decisions with Legal Effects

In accordance with Article 22 of UK GDPR, we do not use AI to make decisions that produce legal effects or similarly significant effects on users without human review. Specifically:

• AI-generated grade predictions do not determine school grades, exam results, or educational placements.
• AI-generated content recommendations do not limit a student's access to any curriculum area.
• Safeguarding alerts trigger human review, not automated punitive action.
• No student is excluded, suspended, or penalised based solely on an AI assessment.

You have the right to request human review of any automated assessment made by the Platform. Contact [email protected].

9. Accuracy and Quality

We take the following steps to maintain the quality and accuracy of AI-generated content:

• Teacher review gate: All lesson content must be approved by a teacher before publication.
• Content quality scoring: AI-generated content is scored against curriculum standards, and low-scoring content is flagged for additional review.
• Feedback mechanisms: Students and teachers can flag inaccurate content through the Platform, which triggers a review.
• Regular audits: We conduct periodic audits of AI-generated content quality and accuracy.
• Model updates: We evaluate and test AI model updates before deploying them to ensure consistent quality.

10. Changes to AI Usage

If we make material changes to how AI is used on the Platform (such as introducing a new AI model, a new AI-powered feature, or a significant change to data processing), we will:

• Update this AI Usage Disclosure
• Notify users and parents by email and through the Platform at least 30 days before the change takes effect
• Where the change involves children's data or new automated processing, conduct a Data Protection Impact Assessment

11. Contact

If you have questions about how AI is used on the Platform, or if you wish to exercise your opt-out rights:

• General enquiries: [email protected]
• Data protection and AI rights: [email protected]
• Safeguarding: [email protected]

This AI Usage Disclosure is effective from 15 February 2026.