Privacy

01Who we are, and which of us is responsible

Outclass Education Technologies Ltd (“Outclass”, “we”, “us”) provides a tutoring and teaching platform. Outclass and OxbridgeAI are trading names of the same company.

Which data-protection role we hold depends on how you reached us, and it matters because it decides who you go to about your data.

• Through your school or tutoring agency. Your school or agency is the data controller. They decide what is collected and why; we act as their processor and only on their instructions. Requests about your data go to them first — they can direct us, and we will act on it.
• Directly, as an individual tutor or student. We are the controller, and you can contact us directly using the details at the end of this page.

02What we collect

• Account details. Name, email address, and a password stored only as a salted scrypt hash — we never hold your actual password. Optionally a profile image, school or organisation, and year of study.
• Learning activity. Questions attempted, answers given, marks, topic mastery, study plans and schedules, and when you were last active. This is the record the platform exists to build.
• Work you submit. Photographs or scans of handwritten working, and typed answers, so they can be transcribed and marked.
• Conversations with our AI. What you type to Marlow and our practice assistants, and what they reply. In school and agency workspaces these transcripts are retained so a teacher, tutor or safeguarding lead can review them — this is a deliberate safeguarding measure, not a hidden one.
• Technical data. A session cookie to keep you signed in, and server logs including IP address for security and troubleshooting.

03What we do not do

• We do not sell your data. Not to anyone, in any form, for any purpose.
• We do not use third-party analytics or advertising trackers. There is no Google Analytics, no advertising pixel, and no third-party tracking script anywhere in this product. Usage statistics are computed inside your own workspace.
• Your data is not used to train AI models. Google contractually commits not to use customer data to train or fine-tune its models without permission, and our handwriting-recognition provider is configured with model improvement switched off. We do not train models on pupil work.

04Where your data is stored

The database holding your records, and the servers running the platform, are located in London, in Google Cloud’s europe-west2 region. The database is reachable only over a private network and only over an encrypted connection.

Each school and each agency gets its own separate physical database. Data belonging to one organisation is not stored alongside another’s.

05Where AI processing happens — stated plainly

This is separate from storage, and we would rather be exact than reassuring.

When the AI marks work or answers a question, the text of that request is sent to Google’s Gemini service for processing. We currently use Google’s global endpoint, which means Google may carry out that processing in a data centre outside the UK. We cannot presently guarantee otherwise: the Gemini model we rely on is not offered in a UK region.

That processing is transient. The request is processed and returned; Google does not retain it as a matter of course, and does not use it for training. Your stored records stay in London throughout.

Google does not retain the requests we send. We hold an approved exception from Google’s prompt-logging for abuse monitoring, so prompts are processed and discarded rather than stored.

06Who else processes your data

These are our sub-processors. Some are only engaged if your organisation switches on the relevant feature; we list them anyway, because “optional” still means your data reaches them when it is on.

• Google Cloud Platform. Hosting, database, and AI processing (Vertex AI / Gemini). United Kingdom for hosting and storage; see the section above for AI processing.
• Mathpix. Transcribing handwritten mathematical work into text, configured so submissions are not used to improve their service.
• OpenAI. Speech synthesis for spoken explanations, and, where enabled, generating search indexes for teaching materials.
• Google Workspace. Sending service email such as sign-in verification and password resets.
• Microsoft. Optional. Exporting lesson plans to OneNote, where a teacher connects their account.
• Telegram, and Meta (WhatsApp). Optional. Delivering messages between a tutor and a parent, where an organisation enables it.
• TutorCruncher. Optional. Synchronising scheduling and roster information for tutoring agencies that use it.
• Stripe. Processing card payments for the direct subscription. Card details are collected and stored by Stripe, not by us.

07How long we keep it

• While your account is open. Your learning record is kept so the platform can track progress over time — that is the point of it.
• When you delete your account. Your personal data is erased, not archived or hidden. That includes your profile, learning history, AI conversation transcripts, and stored preferences.
• When an organisation leaves. Their entire database is deleted, which removes every record belonging to them and their users.
• Parent messaging setup. A pending parent-pairing record is held for at most 75 minutes and then deleted automatically.

08Children and young people

Most of our students are aged 13 to 18, and some are younger. We design for that.

We apply the principles of the Information Commissioner’s Age Appropriate Design Code: we collect only what the service needs, we do not profile children for advertising, we use no engagement-maximising or manipulative design patterns, and we do not build behavioural advertising profiles.

Our AI assistant additionally refuses to store sensitive personal information a student mentions in conversation — including anything relating to health, wellbeing, additional needs, beliefs, identity or safeguarding. Such statements are excluded from the memory the assistant builds.

Where the platform is provided through a school, the school is responsible for the lawful basis for a child’s use of it, including any consent required.

09How we protect it

• Separation. A separate physical database per organisation, with access enforced in code rather than by query filters.
• Encryption. Encrypted in transit, and encrypted at rest by the database. Third-party access tokens are additionally encrypted with AES-256-GCM.
• Passwords. Stored as salted scrypt hashes and compared in constant time. We cannot see or recover your password.
• Sessions. Session cookies are HTTP-only and, in production, restricted to encrypted connections.
• Sign-in. Sign in with Google or Microsoft, so your organisation keeps control of the account.

10Your rights

Under UK data protection law you may ask for a copy of your data, ask us to correct it, ask us to delete it, object to or restrict how it is used, or ask for it in a portable form.

You can delete your account and its data yourself from your account settings at any time. If you use Outclass through a school or agency, direct requests to them, as they are the controller — we support them in responding.

We respond within one month. If you are not satisfied, you can complain to the Information Commissioner’s Office at ico.org.uk.

11Data breaches

If a breach occurs affecting personal data, we notify affected organisations without undue delay and within 72 hours of becoming aware, with what we know, what we are doing, and what they may need to do. Where we act as processor, we notify the controller so they can meet their own reporting duties.

12Changes to this page

When this policy changes materially we update the date at the top and, where the change affects how organisations’ data is handled, tell them directly. We do not make changes retrospectively quiet.