What we use, why we use it, and how long we keep it.

Paper Tiger is operated by Paper Tiger Learning Technologies Ltd, a company registered in England and Wales under company number 17308917. Registered office: 167-169 Great Portland Street, 5th Floor, London, W1W 5PF.

For privacy questions or requests, email hello@paper-tiger.co.uk. This mailbox is monitored.

Accounts are for learners aged 16 or over. Every purchaser must be aged 18 or over. Do not create or use an account if you are under 16.

Paper Tiger asks you to confirm that you are at least 16. We do not ask for your date of birth or exact age, and we apply the relevant privacy safeguards to every learner account. If you are 16 or 17, you cannot currently complete paid checkout. Paper Tiger does not yet offer a separate parent-purchase route.

The learning account belongs to the learner. Paper Tiger would disclose account information only where the learner asks us to, the law requires it, or a serious safeguarding or security situation makes disclosure necessary and lawful.

Optional analytics are off by default. Paper Tiger does not sell or license identifiable learner data, use learner content to train its own or another provider's general-purpose AI models, disclose learner profiles to employers, use behavioural advertising, or run exam proctoring. It may use properly anonymised statistics about broad patterns, but not a learner's transcript, answer, feedback, account or identifiable profile. It will not introduce model training, employer access, advertising or other materially different uses for 16- or 17-year-olds without a new assessment, age-appropriate notice and any legally required choice.

Read the shorter young learner privacy summary, or use the contact details above to ask a question or exercise a privacy right.

Paper Tiger does not sell, license or disclose identifiable or pseudonymised learner data. In particular, it does not provide learner records or progress profiles to employers, institutions or advertisers, and does not use learner content to train general-purpose AI models.

Paper Tiger may create statistics about overall service or learning patterns, such as broad topic-level trends, only after removing or aggregating information so that no individual learner can reasonably be identified. No raw transcript, answer, feedback, account identifier, small-group result or learner profile is included in an external report. No external aggregate report is currently enabled. Separately, a restricted owner-administrator may open an individual learner's account identity and minimum study metadata where reasonably necessary to operate or support the learning service. That audited view includes topic and module references, activity dates and counts, review state, progress status and coarse learning outcomes. It does not load transcript, answer, feedback, prompt, source, flashcard or raw AI content, and it redacts private Studio titles. A separate owner-only quality-control view may be opened to verify the accuracy, relevance and safety of automatically generated learning materials and diagnose generation defects. That separately audited view can show the generated task question, best answer, Quick Check question and answer reference, and flashcard fronts and backs, grouped by building block, synthesis or Mastery Check. It does not retrieve the learner's transcript, submitted answer, learner-specific feedback, uploaded source text or raw AI response. Cross-user quality-control browsing uses a pseudonymous account reference by default. These are core-service administration and quality assurance, not optional product analytics, advertising, employer reporting or exam proctoring.

A learner can export the information Paper Tiger makes available and choose to share that export with someone else. That learner-directed action does not give the recipient ongoing account access or permission to inspect the learner's account. Any materially different use, including model training, employer access, advertising or research reuse, would require a new assessment and updated information before it was introduced.

“Contract” means processing needed to provide the account or feature you request. “Legitimate interests” means a proportionate operational need that we have assessed against your rights. “Legal obligation” means processing needed to comply with law. Consent applies only to optional processing and can be withdrawn.

We receive personal data from:

• you, when you register, study, upload material, buy a pack, change settings, contact support or make a privacy request
• Google or Apple if you choose that sign-in method
• the person or organisation that sends you an invitation, where invited access is used
• your device, browser and our service logs when you use Paper Tiger
• Stripe when a payment succeeds, fails, is refunded or is disputed
• our service providers when they return a transcription, AI result, delivery status or security signal

Please do not submit sensitive information about real people or material you do not have the right to use. Paper Tiger is a study service, not a place to store client, patient, employee or case files.

Grading and learning support

When you ask Paper Tiger to grade a text attempt, we send the transcript, the relevant prompt and answer-key context, and limited technical instructions to Novita AI as the primary text-grading provider. OpenAI receives the same necessary grading content only when the configured quality fallback is used. Diagram grading remains with OpenAI. We use the response to show and save study feedback. The transcript remains editable before grading. AI output may be incomplete or wrong and should be checked as study support.

Novita AI's published API terms state that it does not log, store or retain API user content without prior written consent. See Novita AI's Terms of Service. OpenAI's default API abuse-monitoring logs may retain customer content for up to 30 days unless approved Modified Abuse Monitoring or Zero Data Retention controls apply. Some endpoint-specific and legal or safety exceptions apply. See OpenAI's API data-controls information.

Voice transcription

Paper Tiger holds a recording in memory only while the transcription request is in progress. We do not save the audio. The recording is sent to OpenAI's audio- transcription endpoint. OpenAI's current endpoint table lists no abuse-monitoring or application-state retention for that endpoint, although legal or safety exceptions may apply. The returned text becomes an editable transcript; it is saved only when the study flow saves it. In the free study-pack sample, the transcript of a spoken sample answer is kept with your sample attempt so your marks stay reviewable; the audio itself is never saved there either, and sample content is removed under the retention schedule.

Handwriting

Images are sent to OpenAI only when you request handwriting extraction or a diagram description. Paper Tiger does not save the raw image after the request. Text you choose to place in the composer may later be saved with the attempt. OpenAI's applicable API retention rules still apply to the provider copy.

Studio source files

Paper Tiger extracts text from an uploaded DOCX, text PDF, TXT or Markdown file on the server and does not save the raw file. You can review and trim the extracted text. The raw file is not sent to OpenAI. When you choose “Suggest recall structure”, the reviewed text is sent to OpenAI. A module is saved only after you review and create it.

Essential cookies support sign-in, session refresh and security. The browser cookie that records your analytics choice lasts 12 months. The matching local-storage value remains until you change it or clear browser storage.

Optional analytics are off unless you consent. If you consent, Paper Tiger records coarse route, feature, reliability and learning-flow events for up to 12 months. You can withdraw consent in Settings. We do not use advertising cookies, marketing pixels or session replay.

Paper Tiger uses Supabase for authentication and database services, Vercel for hosting and delivery, Novita AI for primary text grading, OpenAI for fallback grading and other requested AI features, Sentry for scrubbed error and performance reporting, and Stripe for checkout and payment processing. Paper Tiger does not enable Sentry session replay, send default personally identifiable information or intentionally send request bodies, cookies, query strings or non-essential request headers to Sentry. Each provider acts under its own contractual and data-protection terms for the service it supplies.

A provider may process personal data outside the UK. Where a destination is not covered by UK adequacy regulations, an appropriate safeguard is required, such as the UK International Data Transfer Agreement or UK Addendum to approved contractual clauses, together with a transfer risk assessment where required. Email hello@paper-tiger.co.uk to ask for current provider, processing-location or safeguard details.

Paper Tiger uses access controls, row-level database policies, server-side secret handling and audit records. No online service can promise perfect security.

When account deletion completes, ordinary account and study data are deleted and the authentication identity is permanently removed. Necessary legal, privacy, billing and audit evidence may be kept in minimised form for the periods shown above. This means removing direct identifiers and free text where they are no longer necessary.

Depending on the processing and lawful basis, you may ask us to access, correct or erase personal data; restrict how it is used; provide portable data; or consider an objection. Where processing relies on consent, you can withdraw consent at any time without affecting earlier lawful processing. Some rights do not apply in every case, and the law may require or permit us to keep limited records.

Account holders can export data, make privacy requests, and schedule account deletion in Settings. Deletion can be queued as soon as possible or scheduled after a user-selected 28-day recovery window; it can be cancelled until processing starts. Anyone can email hello@paper-tiger.co.uk. We may need to verify your identity. We normally respond within one month; we will explain if the law allows more time or if a request cannot be fulfilled in full.

Please email hello@paper-tiger.co.uk so we can investigate a privacy concern. Contacting us first is not a condition of complaining to the Information Commissioner's Office (ICO); you may contact the ICO at any time. See the ICO's data-protection complaint guidance.

We will update the date and version when this notice changes. If a change materially affects account processing, we will show it to signed-in users and ask for any confirmation the change requires. We do not reset analytics consent unless its purpose or processing changes in a way that requires a new choice.