Safety and privacy

You'll always know it's an AI

Tutortle is an AI tutor, not a person. It says so when a learner first meets it, and again whenever a learner seems unsure. It never claims to have feelings, a body or a life outside the app, and it doesn't present itself as a friend in place of the real people in a learner's life.

Parents and guardians are in charge

• A parent or guardian creates every account for a learner under 18 and gives consent before it starts.
• Each learner gets settings for their age group, which decide what topics and tone are allowed.
• Parents see progress, choose quiet hours and set how often Tutortle may check in.
• Parents can download or delete their child's data at any time.
• There are no ads, no public profiles, and no way for learners to contact other users.

How every message is checked

Safety runs in a separate layer around the tutor. The tutoring helpers can't change its rules or switch it off.

1. Before tutoring. A learner's message is screened for safety concerns, and names, schools and contact details are removed.
2. While tutoring. The tutor works within the topics and tone allowed for the learner's age group.
3. Before a reply is shown. A separate check reads every reply and either lets it through, rewrites it, stops it, or holds it.
4. When something is held. A trained person reviews it before anything is sent.

If a learner is in distress

If a learner mentions self-harm, abuse or being in danger, Tutortle stops tutoring. It shows a short message, written by people rather than generated, with emergency and child-support contacts for the learner's country, and it alerts a member of our team straight away.

Check-ins without pressure

Tutortle's check-ins invite, they don't push. There are no guilt messages about broken streaks, no countdowns and no fake urgency. Nothing is sent during the quiet hours a parent chooses, or beyond the weekly limit they set.

What we collect

• From parents: name, email address and consent records.
• About learners: a first name or nickname, an age range, and the subjects they're studying.
• Learning progress: which topics are solid or shaky, review dates, and answers to exercises.
• Conversations: the messages exchanged with the tutor, kept only as long as described below.

We don't sell personal data, use it for advertising, or share it with anyone except the service providers we need to run Tutortle.

How the AI handles your data

Tutortle's tutoring uses Claude models from Anthropic, through Anthropic's commercial API. Before any text is sent, we remove names, schools and contact details, and learners are identified only by a random ID.

Under Anthropic's commercial terms, data sent through the API isn't used to train Anthropic's models without permission. We also plan to ask Anthropic for zero data retention, so conversation content isn't stored after a reply is returned.

Where data lives and for how long

• Learner records will be encrypted and stored on servers in Türkiye or the European Union.
• Conversations are deleted on a fixed schedule. We keep a compact progress record while the account is open.
• Closing an account deletes its data, apart from anything the law requires us to keep.

Your rights under KVKK and GDPR

We process personal data in line with Türkiye's Personal Data Protection Law (KVKK, Law No. 6698) and, for learners in the European Union, the GDPR. You can ask to see, correct, download or delete your data, and to object to how it's used.

Sending text to our AI provider counts as a transfer of data abroad. We'll use the safeguards Turkish law requires for this, such as standard contracts, and we'll tell families before the beta starts.

This page describes commitments for a product that isn't live yet. A full privacy notice will be published before the closed beta opens.

Contact us

Questions, concerns or a safety report: email [email protected]. We read every message.