Last updated: 8 November 2026 · Effective: [Effective date]

Privacy Policy

What we collect when you learn with Ambrion, why we use it, who helps us, and the choices you have.

Training AI with your learning data is your choice

We use your answers, writing, speaking recordings and transcripts, progress and interactions to provide the AI features. We use them to train and improve AI models only if you turn this on; it is off unless you do. See section 4.

Each section starts with a short summary, followed by the full text.

  1. Who we are
  2. The data we collect
  3. How we use your data, and why
  4. How we use your data to train and improve AI
  5. Who we share it with, and where
  6. How long we keep it
  7. How we protect it
  8. Your rights and choices
  9. Children
  10. Cookies and local storage
  11. Changes to this policy
  12. Contact

1. Who we are

In short

[Legal entity name] runs Ambrion and is responsible for your personal data.

This Privacy Policy explains how [Legal entity name] (“Ambrion”, “we”, “us”), of [Address], collects and uses personal data when you use the Ambrion website, web app and phone apps (the “Service”). We are the controller of that data. You can reach us about privacy at [Contact email].

Please read this policy together with our Terms of Service.

2. The data we collect

In short

Your account details, everything you do while learning (including what you write and record), what you send us, and basic technical data. Card payments are handled by Stripe, so we never see your full card number.

Account details

  • your name and email address, and your password (stored only in a protected, hashed form by our sign-in provider);
  • if you continue with Google: the name, email address and profile picture Google shares with us. We never see your Google password;
  • the date you agreed to the Terms and this policy, and which version you agreed to.

Learning data

  • Answers: what you choose, type or skip in exercises, tests and cards, and the results;
  • Writing: the texts you write, drafts, and the corrections and feedback you receive;
  • Speaking: the audio of your recordings and their transcripts, with feedback and scores;
  • Progress: scores, level estimates, the words you know or save, streaks, goals, the languages you learn, and exam dates or targets you tell us;
  • Interactions: what you open, how long you spend, what you replay or skip, and the order you do things in.

What you send us

Messages, feedback and any screenshots or files you attach when you contact us or report a problem.

Technical data

Your IP address, browser and device type, the pages you request, timestamps, and error reports (which can include your account ID and the page you were on). We use this to run, secure and fix the Service.

Payments

When you subscribe on the website, Stripe, our payment provider, collects your card details and processes the payment. We never see or store your full card number. We receive what we need to run your subscription: your plan, its status and dates, the payment amounts, and a fingerprint of the card (a code that identifies the card without revealing its number), which we use to offer the free trial once per card. When you subscribe on iPhone, Apple processes the payment and tells us the state of your subscription.

Please do not include sensitive information (such as health details, or other people’s personal data) in what you write or record. Your voice recordings are personal data; we do not use them to identify you by your voice.

3. How we use your data, and why

In short

To run Ambrion for you and power its AI, to train and improve that AI only if you turn it on, to keep it safe, and to meet legal duties.

We use personal data for these purposes, on these legal bases:

  • Providing the Service: creating your account, signing you in, saving your progress, giving feedback and scores, and choosing your next practice. Basis: performing our contract with you (the Terms).
  • Training and improving AI: see section 4. Basis: your consent, which is optional and which you give or withdraw under Settings › Privacy & AI on the Ambrion website.
  • Improving the Service: understanding which exercises help, fixing problems, and building new features. Basis: our legitimate interests.
  • Safety and security: preventing abuse, fraud and attacks, and enforcing the Terms. Basis: our legitimate interests and legal obligations.
  • Talking to you: answering your messages and sending important notices about your account or these documents. Basis: performing our contract and our legitimate interests.
  • Legal duties: keeping records and answering lawful requests. Basis: legal obligations.

We do not sell your personal data, and we do not use it for advertising. We do not make decisions about you that have legal or similarly significant effects based solely on automated processing; AI scores and level estimates in the Service are practice estimates only.

4. How we use your data to train and improve AI

In short

Your learning data is used to run the AI features. It is used to train, test and improve AI models only if you agree. It is off unless you turn it on, and you can turn it off at any time.

What we use

All of your learning data: your answers, your writing, your speaking recordings and transcripts, your progress, and your interactions with the Service, together with the feedback and scores the Service gave you.

What we use it for

  • To provide the AI features to you. Your writing, transcripts and answers are sent to AI models to produce feedback, corrections, scores, level estimates and your next practice. Your recordings are sent to a speech-to-text model to produce transcripts.
  • To train and improve AI models, if you agree. We use learning data to train, fine-tune and evaluate the models behind Ambrion (for example models that check writing and speaking, estimate levels, transcribe learner speech, and pick what to practise next), to build test sets that measure their quality, and to improve our prompts and content. This includes our own models and models we build or tune with our AI providers.
  • Human review. Members of our team, or people working for us under confidentiality, may review samples of learning data and AI output to check and improve quality.

How we protect it

Before learning data is used to train or evaluate models, we remove direct identifiers such as your name and email address and replace your account ID with a random code. Voice recordings cannot be made fully anonymous, because a voice is personal, so we keep them to the smallest group of people and systems that need them. We do not sell your data, and we do not publish it.

It is your choice

Using your learning data to train and improve AI is optional and off by default. Agreeing to these terms when you create your account does not turn it on: you choose it separately under Settings › Privacy & AI on the Ambrion website, and the choice applies to your whole account, including the iPhone app. Saying no does not change what you can do in Ambrion. Running the AI features for you, described above, is part of the Service and is not affected by this choice.

If you change your mind

You can withdraw your agreement at any time under Settings › Privacy & AI on the website, or by closing your account (see section 8). From then on we stop using your data for new training; if you close your account, we delete it as described in section 6. AI models already trained or tuned using de-identified data cannot un-learn it, so they may continue to be used. Withdrawing does not affect use that happened before.

5. Who we share it with, and where

In short

Only with companies that help us run Ambrion, under contract, and when the law requires. Some of them are outside your country.

We share personal data with service providers who process it on our behalf and only on our instructions. Today they are:

  • Database, sign-in and file storage: Supabase (your account, learning data, recordings and files you send us). Sign-in emails are sent through it too.
  • Payments on the website: Stripe (your email address, your card details, which Stripe collects itself, and your subscription and payments).
  • Hosting: Vercel (the website and web app) and Fly.io (our servers and background jobs).
  • AI feedback on writing and speaking: OpenAI (what you write, the transcripts of your recordings and the task, to check them and give feedback and scores).
  • Speech-to-text: Groq (your speaking recordings, to transcribe them).
  • Feedback in the older practice exercises, practice material and computer voices: Google, through the Gemini API (your answers in those exercises, to give feedback; practice material we write).
  • Deep Read: Google, through the Gemini API (the text you paste into Deep Read, to build the reading lesson from it). It is sent only when you ask for a lesson, so paste only text you are happy to share.
  • Computer voices: ElevenLabs (audio for practice material we write; we do not send it your recordings).
  • Writing and checking practice material: OpenAI and DeepSeek (practice texts and questions prepared in advance, apart from the feedback above).

Your choice about AI feedback

Before your first writing or speaking card, we ask whether we may send your answers and recordings to these AI services for feedback. If you say no, we do not send them: your writing and speaking cards are still saved with us, with no feedback, and the older writing and speaking exercises, which need these services, are not available. You can change your answer at any time under Settings › Privacy & AI; turning AI feedback on also sends the writing and speaking answers you saved in the last 7 days. This choice is separate from the use of learning data to train and improve AI described in section 4.

If you choose to continue with Google, Google also processes your sign-in under its own privacy policy.

We may also share data if the law requires it, to protect people’s rights and safety, or as part of a merger, sale or reorganisation of our business, in which case this policy will continue to protect your data.

International transfers

Our providers process data in the United States and other countries, which may have different data protection laws from yours. Where the law requires it, we protect these transfers with safeguards such as the European Commission’s standard contractual clauses or an adequacy decision.

6. How long we keep it

In short

For as long as your account is open. When you close it, we delete your data, apart from what the law makes us keep.

We keep your personal data for as long as your account is open, because your history is what the Service uses to plan your practice.

When you close your account, we delete your account, learning data, recordings and files from our live systems, normally straight away and at the latest within 30 days. Copies in backups are deleted as the backups expire. We keep data longer only where the law requires it, or to deal with a dispute or abuse.

De-identified data that has already been used to train or evaluate AI models, and the models themselves, may be kept after your account is closed (see section 4).

Server logs and error reports are kept for a limited time and then deleted.

7. How we protect it

In short

We use encryption, access controls and private storage, and limit who can see your data.

Data is encrypted in transit. Your learning data is protected by access rules in our database so that your account can reach only its own data, recordings are kept in private storage, and only people who need access to run the Service have it. No system is perfectly secure; if a breach affects your data, we will tell you and the authorities as the law requires.

8. Your rights and choices

In short

You can ask to see, correct, export or delete your data, and turn AI training and AI feedback on or off at any time.

Depending on where you live, you have the right to:

  • Access the personal data we hold about you and get a copy;
  • Correct data that is wrong or incomplete;
  • Delete your data, by closing your account;
  • Export your data in a portable, machine-readable format;
  • Object to or ask us to restrict some uses of your data;
  • Turn AI feedback on or off at any time under Settings › Privacy & AI (see who we share it with);
  • Withdraw consent to the use of your learning data to train and improve AI, at any time under Settings › Privacy & AI on the Ambrion website. Withdrawing does not affect what happened before.

To use any of these rights, write to us at [Contact email] from the email address on your account, or use the account tools in the app where they are available. We may need to confirm it is you. We will answer within the time the law sets, usually one month.

You also have the right to complain to your local data protection authority. We would appreciate the chance to help first.

9. Children

In short

Ambrion is for people aged 16 and over.

The Service is not meant for anyone under 16, or under the higher age of digital consent where they live. We do not knowingly collect data from them. If you believe a child has created an account, contact us at [Contact email] and we will delete it.

10. Cookies and local storage

In short

We use only what is needed to sign you in and remember your settings. No advertising or tracking cookies.

We use a small number of cookies and similar browser storage, all needed for the Service to work:

  • Sign-in cookies that keep you signed in securely;
  • A language cookie that remembers which language you are learning;
  • Local storage in your browser for settings (such as playback speed), unsent drafts and answers, and practice saved for offline use.

We do not use advertising cookies or third-party analytics cookies. You can clear cookies and storage in your browser settings, but you will be signed out and some features may stop working.

11. Changes to this policy

In short

We will tell you about important changes before they apply.

We may update this policy as the Service or the law changes. The date at the top shows when it was last updated. If a change is material, especially to how we use your data for AI, we will tell you in advance by email or in the Service, and where the law requires it we will ask you to agree again.

12. Contact

In short

Privacy questions and requests go to [Contact email].

[Legal entity name], [Address]. Email: [Contact email]. See also our Terms of Service.