Privacy Policy

• We collect what we need to run your courses: account details, your course setup, the files you upload, the text we extract from them, your answers and progress, and technical logs.
• Your prompts, the excerpts we retrieve from your uploads, and your answers are sent to our AI provider so a lesson or reply can be generated.
• We do not use your content to train ClassDay models or any third-party foundation model.
• We do not sell personal information and we do not share it for cross-context behavioural advertising.
• No analytics, advertising or tracking tools are installed. The only browser storage we set is what signing in and remembering interface preferences require.
• Deleting a course deletes its files, extracted text, lessons and progress. Account deletion is a verified request you can start in Settings.

This policy covers the ClassDay web application and consumer accounts. The controller (or "business") is to be confirmed, to be confirmed. Privacy contact: to be confirmed.

The table below maps every category we actually store, with real examples from the product. It is the authoritative summary; the sections that follow add detail.

• From you: registration, onboarding, course setup, uploads, tutor questions, written answers, requests and reports.
• Automatically: request metadata such as IP address and user agent, rate-limit counters, usage and token logs, error logs.
• Derived by us: extracted document text, indexed chunks, summaries, mastery and confusion signals, generated lessons and schedules.
• From a sign-in provider: if you sign in with Google, the identifier and email that provider returns.

We do not buy personal information from data brokers and we do not enrich your profile from third-party sources.

• Provide the service: parse a syllabus, design a curriculum, build a schedule, generate and run lessons, evaluate answers, track progress.
• Keep the tutor scoped to your course and decline unsafe or out-of-scope requests.
• Enforce plan allowances, rate limits and concurrency caps, and meter AI cost.
• Secure the service, investigate abuse and prevent fraud.
• Support you and respond to privacy, content and copyright requests.
• Diagnose failures, for example a lesson generation that times out.
• Comply with legal obligations and enforce our Terms.

We do not use your data for profiling that produces legal or similarly significant effects, and we do not make automated decisions of that kind about you.

• Contract: account creation, course generation, teaching, progress, billing plan administration.
• Legitimate interests: security, abuse and rate-limit enforcement, debugging, service improvement that does not involve model training, and defending legal claims.
• Consent: optional marketing email, and any optional improvement permission we may offer. You can withdraw consent at any time.
• Legal obligation: responding to lawful requests and keeping required records.

Please do not upload special-category data (for example health or biometric information) or other people's sensitive data. We do not ask for it and we do not need it to teach you.

When you generate a curriculum, open a class, answer a question or ask the tutor, our server assembles a prompt from your course setup, relevant retrieved excerpts of your uploaded materials, your question or answer, and our teaching instructions. That prompt is sent through the Lovable AI Gateway to the model provider, which returns generated text. We store the output in your account and record token counts and cost units. We do not send your email address, password or payment details to the model provider.

Whole files are not shipped wholesale for every request: retrieval selects passages. However, a syllabus analysis or document summary can include a large portion of a single document's extracted text.

For counsel: Confirm and record the model provider's data-use, retention and zero-retention terms as they apply through the gateway, and reflect anything that differs from the statement above before publishing.

Current setting: no training. Your uploads, prompts, answers and generated content are not used to train, fine-tune or evaluate ClassDay models, and we do not permit their use to train third-party foundation models. Our requests are sent with provider storage disabled. If this ever changes, we will publish the change, update this policy, and, where the change involves your content, ask for your consent rather than assume it.

We use a small set of providers to run the service. Each processes data on our instructions for the purpose listed. The current list, with data categories, is on the Subprocessors page. Today that is our application platform, our managed database, authentication and storage provider, the AI gateway and the model provider behind it.

Categories we do not currently use: payment processing, product analytics, advertising or marketing platforms, third-party error monitoring, and standalone support helpdesk software.

• Legal process and safety: when required by law or valid legal process, or where we believe in good faith that disclosure is necessary to prevent serious harm, fraud or a security threat.
• Corporate transactions: in a merger, acquisition, financing or asset sale, subject to this policy continuing to apply or you being notified of a material change.
• At your direction: when you ask us to share something, for example by exporting it.

We do not sell personal information, we do not share it for cross-context behavioural advertising, and we do not use or disclose sensitive personal information for purposes that require an opt-out under US state law. This has been true for the whole life of the product, including the preceding twelve months.

We use only what the product needs to work. Details are in the Cookie Notice. In short: your signed-in session is stored in your browser's local storage by the authentication library, and a small first-party cookie remembers whether the app sidebar is open. There are no analytics, advertising or cross-site tracking technologies, so there is nothing to opt out of today.

We keep data against criteria rather than invented fixed periods:

• Course content, uploads, extracted text, lessons and progress: until you delete the course or your account. Course deletion removes the database records and the stored files immediately.
• Account and profile: for the life of the account.
• Rate-limit events: purged automatically; the current job removes rows older than about two hours.
• AI usage and cost logs: kept while needed for plan accounting, dispute handling and abuse investigation, then removed.
• Policy acceptance records: kept for as long as the agreement could matter, because they evidence what you agreed to.
• Age attestation records: the statement text you confirmed, its version, the timestamp and a hashed IP address, kept while the account exists as evidence of the age check.
• Reports of an unauthorised child account: kept as a record of the report and the action taken, with only what is legally or security-necessary preserved from a suspended account.
• Requests, reports and support mail: kept as a record of how we handled them.
• Backups: our infrastructure provider keeps routine backups. Deleted data can persist in a backup until that backup ages out on its normal schedule, and is not restored into the live service.

For counsel: Confirm the backup retention window offered by the infrastructure plan and, if a specific period is required by any applicable law, publish it here once verified.

Signed in, from Settings → Privacy & data, you can:

• edit your profile details directly;
• delete an uploaded material or an entire course, which also deletes the stored files;
• request a copy of your data, and follow the request's status;
• request account deletion, with confirmation, and follow its status;
• set your marketing email preference.

Because these requests concern your own data, we act on them from your authenticated session and may ask for re-verification before completing a deletion. We do not claim instant erasure: requests are tracked with a status, and residual copies in backups age out on their own schedule.

Uploaded files are held in a private storage bucket that is not publicly readable. Every table storing your data has row-level security so records are readable only by the account that owns them, with server-side ownership checks in addition. Server functions that touch your data require an authenticated session, and requests are protected against cross-site invocation. Privileged operations are separated from ordinary reads.

No system is perfectly secure, and we cannot guarantee that your data will never be accessed improperly. We do not claim any certification or audit (such as SOC 2 or ISO 27001), and none should be inferred.

Our providers may process data in countries other than yours, including the United States. Where a transfer from the EEA, UK or Switzerland requires a safeguard, we rely on the mechanism offered by the relevant provider. Confirmed processing locations and transfer mechanisms are published on the Subprocessors page as each is verified.

ClassDay is not available at all to anyone under 16. At signup you tick a single checkbox confirming: "I confirm that I am at least 16 years old." We do not ask for a date of birth, birth year or any other age detail. We record the attestation text, its version, a timestamp and a hashed IP address in public.age_attestations, and mirror it into your policy acceptance record, as evidence of the check.

Users aged 16 up to the age of majority where they live may use ClassDay only with the permission of a parent or legal guardian where applicable law requires it. There is no separate profile or account type for this; the user holds and uses their own account.

ClassDay is not directed to anyone under 16, including children under 13, and we do not knowingly collect personal data from them. User content and account data are never used for behavioural advertising, and we do not sell or share any data for cross-context behavioural advertising. We do not use any user's prompts, uploads, answers or progress to train models.

Underage accounts. If we obtain actual knowledge that someone under 16 created or controls an account, we suspend it, preserve only what is legally or security-necessary, verify the matter with the reporting parent or guardian, and delete the account and its data as required by law. A parent or guardian can report an account to our child-safety contact, which is being set up before launch and will be published here: to be confirmed. We keep those reports and their outcome as a record of the action taken.

A consumer ClassDay account is not a school-authorised service. We are not a "school official" with a legitimate educational interest under FERPA, we do not receive education records from an institution, and we do not operate under a school's COPPA consent. If you upload material that your institution treats as an education record, you are doing so as an individual and remain responsible for whether you are permitted to.

Institutions that want us to process student data need a separate written agreement and a data-processing agreement. Contact to be confirmed.

Depending on your state, you may have rights to know or access, delete, correct, obtain a portable copy, and to opt out of sale, sharing for targeted advertising, or certain profiling. As stated above, we do not sell or share for targeted advertising, so there is nothing to opt out of; we honour universal opt-out signals such as Global Privacy Control where they apply. We will not discriminate against you for exercising a right.

Categories of personal information collected, mapped to California's statutory categories: identifiers; customer records information; internet or network activity; commercial information about your plan; professional or education-related information; and inferences drawn about your understanding of course topics. Sources, purposes, disclosure recipients and retention criteria for each are set out above.

For counsel: Confirm which state statutes actually apply given revenue and consumer thresholds, then finalise the notice content, request-verification standard, response deadlines, and any required "Do Not Sell or Share" link (not required today because no sale or sharing occurs).

Where the GDPR or UK GDPR applies you may request access, rectification, erasure, restriction, portability, and object to processing based on legitimate interests; you may withdraw consent at any time; and you may complain to your supervisory authority. In Canada, you may access and correct your personal information and complain to the Office of the Privacy Commissioner.

EU representative: to be confirmed. UK representative: to be confirmed.

For counsel: Determine whether an Article 27 representative and/or a DPO are required before offering the service in those markets.

If we decline a request, we will explain why and you may appeal by replying to our decision or writing to to be confirmed. We will respond to an appeal within the period required by the law that applies to you. An authorised agent may submit a request on your behalf where the law allows; we will ask for proof of authorisation and may ask you to confirm directly.

The fastest route is Settings → Privacy & data while signed in, which creates a tracked request. Otherwise write to to be confirmed or to be confirmed. We may need to verify your identity before acting.

We will update this policy as the product changes and will revise the version and effective date at the top. Material changes get notice in the product, and where legally appropriate we will ask for fresh consent rather than rely on continued use.