Privacy & Data Safety
Language learning should be fun, and we want you to be able to use Geeks in Kyoto by initiating dialogue in your words, on topics and situations relevant to you. At the same time, learning is a semi-public activity. Don't enter deeply personal, sensitive, or confidential details in the chat — just as you wouldn't talk about them in class where they'd be heard by the teacher and other students.
The short version
We send your dialogue to Gemini (Google's AI) to make the tutor reply. We don't store it and we don't look at it.
If you sign in, we keep your account and your settings, and we record how you use the course — which material you open, how long you listen, how often you use the tutor — so we can show you your own stats, tune the course, and keep track of your plan's allowance. We record which material you use, not the words you type or hear. It's stored under a one-way tag: enough for us to show you your own numbers and to delete them with your account, but not tied to what you say. It's part of how the course runs, so there's no separate off-switch. Your choice is to browse without an account or to delete your account, which erases it.
If you're not signed in we still have to count what we hand out, so the free allowance means something. That count is kept against a scrambled form of your network address rather than an account, for a few days.
AI can make mistakes or occasionally say something inappropriate. If you see wrong or inappropriate content, please flag it (menu → Flag content) and a human will review it.
This course is aimed at adults. There are a few frank words about body parts and functions, but nothing violent or sexual.
You can delete your data at any time. If you've entered something you'd rather keep private, email [email protected] and we'll remove it.
What we store
The table is the policy. Everything we keep, why, on what legal basis, and for how long:
| Data | Purpose | Legal basis | Kept for |
|---|---|---|---|
| Account identity — email, name, avatar from Google Sign-In | Sign you in; keep your account yours | Contract | Life of your account; deleted within 30 days of account deletion |
| Settings — your chosen level, register, display and playback preferences | Make your settings stick | Contract | Life of your account |
| Usage statistics — which material you open and finish, its difficulty, and how many minutes you listen; for each AI call, the model, token counts and timing (not the text); a random per-installation identifier. Stored under a one-way tag derived from your account, in a database that holds no names or emails. Records which material, not the words you type or hear | Show you your own stats; tune the course | Contract | Detailed rows pruned after 30 days; your own daily summaries for the life of your account; anonymous totals kept indefinitely |
| Plan allowance — how much of your listening allowance you have drawn, and which items have been handed to you. Kept against your one-way account tag when you're signed in, or against a scrambled form of your network address when you're not | Make the free allowance mean something; unlock what you've paid for | Contract | A few days for the running balance; up to 30 days for the record of what was handed out |
| Abuse limits — a scrambled form of your network address with a count of AI requests that day | Stop one visitor exhausting the AI budget for everyone | Legitimate interest | Days, not months |
| Tutor messages — what you type, sent to Google (Gemini) to generate the reply | Generate the tutor's reply | Contract | Not stored. Processed and discarded |
| Announcement email — only if you tick the box | Be notified by email when we have something big to announce, like going live | Consent (revocable) | Until you unsubscribe |
| Developer updates — only if you tick the box | Hear from the developer occasionally: feature updates, requests for feedback | Consent (revocable) | Until you untick it |
| Network & hosting (Cloudflare) — handles your network address and approximate location in order to serve and protect the site, and gives us cookieless, aggregate traffic figures | Keep the service up; stop attacks; understand overall traffic | Legitimate interest | Cloudflare's own retention; the only piece we keep in our own statistics is a coarse country |
| Reported AI responses — the response, surrounding context, your account, your reason, when you report one | Human review of flagged AI output | Legitimate interest; platform policy | Up to 24 months |
One thing we plan to add will store more: your subscription state, when paid plans open. That isn't collected today. When it turns on, we'll update this policy and tell you in the app first.
What we deliberately do not keep
- Chat transcripts. Your conversation history is not stored.
- A profile built from your words. We work from counts and minutes, not the content of what you typed.
- Your chosen level. We note the difficulty of the material you use, not the level setting you picked.
- Your network address in the clear. Where we have to count something against a visitor who isn't signed in, we scramble it first and keep it briefly.
- Advertising data or trackers. None, first- or third-party.
- Card or billing details. The store you buy through handles those (see Billing).
Who we share your data with
We share your data with technology partners to enable us to provide the service:
- Cloud infrastructure and hosting: Cloudflare
- AI inference: Google (Gemini API)
- Sign-in: Google (Sign in with Google)
- Payment processing: on Android, purchases are processed by Google Play, which acts as Merchant of Record; for purchases made on the website, billing is handled by a third-party payment processor acting as Merchant of Record (named here once finalised). In every case we do not receive or store your card or billing details.
Each has their own privacy policy regarding data held on their systems on our behalf. If we change technology partners, your data will be moved accordingly and this policy will be updated.
Your rights
Depending on where you live, you have rights under laws such as the UK GDPR, EU GDPR, and California Consumer Privacy Act. These include the right to:
- Access the personal data we hold about you
- Have inaccurate data corrected
- Request deletion of your data
- Export your data in a portable format
- Object to or restrict certain processing
To exercise any of these rights, email [email protected]. We will respond within 30 days.
What we won't do
- We will not use your data for advertising or unsolicited marketing purposes.
- We will not share or sell your data to third parties for any purpose other than billing and providing the language learning service itself.
- The two exceptions the law obliges us to name: we may disclose data if legally compelled, and if Geeks in Cities Ltd is acquired your data moves with the service under the same commitments — we would notify you.
Deleting your data
To delete your data, write to [email protected], or delete your account in the app. We remove your data — account, settings, and your account-tagged usage rows — from our live systems within 30 days.
Your data may still exist in archival form such as encrypted site backups, as it is impractical to selectively delete individual records from these. Archived data is retained only for disaster recovery and is not accessed for any other purpose. Backups are rotated out within a reasonable period.
Staff access
Staff and contractors of Geeks in Cities Ltd will technically have access to and the ability to view your learning data, as system administrators routinely do. However, we are bound by professional ethics. Our staff and contractors are prohibited from looking at your learning data or anything you type, except as required to resolve a technical issue you have reported.
AI and your data
Our service is powered by AI. A highly capable AI system reads your chat and acts as your personal tutor. If you have a privacy concern regarding the AI itself reading your chat, this is the wrong service for you.
Billing
We are not yet accepting subscribers. When paid plans go live, billing will be handled by the platform you purchase through, acting as Merchant of Record: Google Play for purchases in the Android app, and a third-party payment processor (named here once finalised) for purchases on the website. We will not store your credit card or billing data, and a link to each processor's privacy policy will be provided here.
Security incidents
If we find that our servers have been breached, we will:
- Make a public announcement.
- To the extent possible, email affected users telling you if we suspect your data has been leaked.
- Sign you out everywhere and require a fresh Google sign-in.
- Show you the options to delete your data or close your account.
Your responsibility
We disclaim liability for harms you might incur if you share highly personal, identifying, confidential, legally challenging, restricted, or otherwise sensitive information with our service. Do not do this. You are responsible for keeping your communication with the service at a public level, as you would in a language class with a teacher and other students.