Effective date: 2026-10-17
Operator: Zhang Yuanyi (individual developer, located in Japan; “we” or “us” in this document)
Service endpoints:
| Endpoint | Address |
|---|---|
| Website | tomoshi.app |
| Dictionary site | dict.tomoshi.app |
| Gateway | api.tomoshi.app |
| Data distribution | data.tomoshi.app |
| Speech content distribution | tts.tomoshi.app |
Contact: [email protected]
Tomoshi is a Japanese-learning dictionary application, available as a desktop version, a mobile version and a companion browser extension (together, the “App”). This policy explains what data the App handles, what data leaves your device, who it is sent to, how long it is kept, and how you control it.
The data-handling rules are the same across all three, with differences only in individual features that are platform-specific — the global hotkey lookup and the browser extension are desktop-only, system backup is mobile-only, and updates to the application itself are handled by the app store on mobile. Every platform-specific point is marked as such below. Data handling for the website (tomoshi.app) is covered separately in Section 7.
Our guiding principle: local first, data minimisation. The App does not require an account — signing in is optional and serves only wordbook cloud sync (see 2.8 and 2.9). The App serves no advertising, embeds no third-party analytics tools (such as Google Analytics), and does not sell or share your data with third parties for marketing.
1. Data That Stays on Your Device
The following is stored locally on your device and is not uploaded to us (the wordbook excepted once you sign in and thereby turn on cloud sync — see 2.9):
- Dictionary lookups and browsing: lookups are performed against the local dictionary database, and what you look up is not sent to any server (except when you trigger a cloud feature — see Section 2, where each case is stated explicitly);
- Search history: always local only, never uploaded by any feature (including cloud sync, see 2.9);
- Wordbook, review progress and reading position: local only by default; only after you sign in, and thereby turn on cloud sync, is it synchronised to the server of your service region as described in 2.9;
- App settings;
- Clipboard reads (desktop only): the global hotkey lookup reads clipboard text to perform a local lookup; that text is not uploaded automatically;
- Browser extension (desktop only): the extension communicates only with the Tomoshi desktop application running on your own device (a loopback address) and does not connect directly to any external server.
Uninstalling the App, or deleting within it (history and wordbook support both per-item and bulk clearing), deletes this data completely, and we cannot recover it — because we never had it.
System backup (mobile only): the system backup features of Android and iOS copy the settings and study data above (wordbook, history, review progress) to your own cloud backup space (Google Drive / iCloud) so you can restore it on a new device. That is a relationship between you and your device vendor; we neither handle nor can access it. Dictionary data (which can simply be downloaded again) and the anonymous device identifier (see 2.1) are explicitly excluded from backup. Whether system backup is enabled is controlled by you in your device’s system settings.
Automatic wordbook backup (iOS only): separately from system backup, the iOS version saves a copy of your wordbook (including review progress, but not your search history) once a day into this app’s folder in your own iCloud Drive (Files app → iCloud Drive → Tomoshi, keeping the last 7 days), so that you can import and restore it manually if you delete and reinstall the App. These files likewise move only between your device and your iCloud account; we neither handle nor can access them. The feature disables itself automatically if you are not signed in to iCloud, or have turned iCloud Drive off for this app. You can delete these backup files at any time in the Files app.
2. Networked Features and the Data They Send
The App’s cloud features are provided through our own gateway server
(api.tomoshi.app). What each feature sends, why, and how long it is kept is
set out below.
2.1 Anonymous Device Identifier
The first time you use a cloud feature, the App registers a randomly generated anonymous device identifier with the server, used for managing daily usage quotas and for abuse prevention. It does not contain, and cannot be used to derive, your name, email address or device hardware information; we have no means of linking it to any real-world identity.
2.2 Cloud Speech Synthesis (Pronunciation Playback)
When you tap the pronunciation button, the word or example sentence being
read aloud and the selected voice are sent to the server for speech
synthesis. Speech is synthesised by Microsoft Azure AI Speech (see Section 5).
Results are cached by text content for reuse, and the cache is not associated
with the device identifier — it is not possible to work backwards from the
cache to “who listened to what”. The audio file itself may be served from a
nearby content delivery network node (tts.tomoshi.app); that request
identifies only the audio content requested and likewise carries no device
identifier.
2.3 AI Analysis
In the following cases, the text you look up (a word or a sentence) is sent to the server, where a large language model generates an analysis:
- You actively tap “AI analysis” or a similar button;
- When a search term you have typed does not match the local dictionary, the App automatically initiates an AI analysis (with the exception of clipboard text captured via the desktop global hotkey — that is sent only if you tap manually).
Model inference is performed by a third-party large-model service (currently OpenAI’s official API by direct connection; requests from mainland China, Hong Kong and Macau are instead handled by a direct connection to DeepSeek’s official API, as OpenAI does not offer its service in those regions. The region a request comes from is determined by the server from the connection, at country/region level (as in 2.6); see Section 5). To avoid regenerating the same content, analysis results enter a shared cache for around 90 days: the cache is keyed on the query text itself and is not associated with the device identifier; a query that hits the cache is not sent to the model provider again.
Active sharing: when you explicitly tap the share button in exchange for quota, the authorisation record for that analysis is retained in association with your anonymous device identifier (for crediting and withdrawal). This is the only case in which “content” is linked to “device”, it happens only when you actively do it, and it can be withdrawn in the App.
2.4 User Feedback (Submitted by You)
Content submitted through the in-app feedback entry point (the dictionary entry reported, the description you write) is sent to us together with the app version, operating system type and anonymous device identifier, and is used to handle your feedback. It is deleted automatically after at most 180 days.
2.5 Crash and Error Reports (Automatic, Can Be Turned Off)
So that defects can be found and fixed, the App automatically sends us a technical report when it crashes or hits an internal error.
- What is sent: the program error message and call stack, where the error occurred (source file/line number), thread name, time of occurrence, app version, operating system type (such as windows/macos/android), and the anonymous device identifier (see 2.1).
- What is explicitly not included: any word or sentence you looked up, clipboard content, or any input or study data such as your wordbook or history; nor any identifying information such as name, email address, IP address or hardware serial number.
- When it is sent: on a crash, the report is first saved locally and sent at the next launch; interface-level internal errors are sent in real time. Each session applies de-duplication and a count limit, so reports are not sent repeatedly.
- Recipient: sent only to our own server (
api.tomoshi.app), never through any third-party error-collection service (such as Sentry). - Retention: deleted automatically after at most 180 days.
- How to turn it off: Settings → General → Diagnostics → turn off “Automatically send crash reports”. This takes effect immediately, and any unsent reports held locally are discarded. Turning it off does not affect any functionality.
2.6 Server Logs and Usage Statistics
Our servers record technical metadata for troubleshooting and capacity planning: the API endpoint requested, time, status code, latency, cache-hit and rate-limit outcomes, model tier, coarse region (country/region level), and the anonymous device identifier. These logs and statistics never contain the content of your lookups — this is an architectural constraint, not after-the-fact filtering. Per-request logs are kept for around 7 days and aggregated statistics for around 3 months.
As with any internet service, our servers encounter your IP address while processing requests, used for rate limiting and abuse prevention during registration; the country/region-level coarse region mentioned above is also derived from it (performed by Cloudflare at the edge, see Section 5) for capacity planning and routing to a nearby endpoint. The IP address itself is not written to logs or statistics, is not associated with lookup content, and is not used to identify or track individuals — all that is retained is an attribution coarse enough to be at country/region level, from which no one can be located or identified.
2.7 Update Checks
The App periodically requests a static manifest file from our distribution
server (data.tomoshi.app) to check whether a new version of the dictionary
data pack (and, on desktop, of the application itself) is available. This is an
ordinary file download request and sends no anonymous device identifier, no
lookup content and no study data; as with any network request, the server
encounters your IP address while handling it (as in 2.6). On mobile, updates to
the application itself are handled by the app store, and the App checks only
for dictionary data packs.
2.8 Account and Email Address (wordbook cloud sync only)
Every feature of the App works without an account. Only if you want your wordbook to stay in sync across several devices do you need to sign in by email verification code — this is the only situation in which the App collects an email address, and it happens only when you actively tap “Sign in with email” and enter your address.
- What is collected: the email address you enter. The verification email is sent by our server through Cloudflare’s email sending service (see Section 5) and contains nothing but the code; no marketing email is ever sent.
- Purpose: solely to verify you at sign-in, recognise your account (one address corresponds to one account) and link the wordbooks on your devices to the same cloud copy (2.9). It is not linked to the content of your lookups — server logs and usage statistics contain no email addresses (as in 2.6).
- Storage: the address is stored encrypted on the server; routine operation uses only an irreversible hash of it to look up the account. The address is not stored on your device; the App shows only a masked form (such as so***@example.com).
- Signed-in state: after signing in, the anonymous device identifier described in 2.1 is associated with your account for 30 days and renewed automatically. “Sign out” removes that association on this device and stops syncing; your local wordbook is kept.
- Dates of use: with your account, the server records the two most recent dates on which you used the service (the date only, in UTC, with no time of day). They are updated when you sync your wordbook while signed in, at most once per account per day. They serve only to count how many accounts use the service each day — a single aggregate number. They are not linked to the content of your lookups and are not used to analyse your individual habits, for profiling or for recommendations; the wordbook and review progress described in 2.9 play no part in this count, and server logs contain no account identifier. They are part of the account record and are deleted with it when you delete your account.
- Retention and deletion: the email address and account record are kept until you delete your account. You can delete it at any time in the App under “Me / Settings → Account → Delete account”, and deletion takes effect immediately: the server removes your email address, account record and wordbook cloud copy (2.9) at once and they cannot be recovered; your local wordbook, history and other data are unaffected. Signing in again creates a new account.
2.9 Wordbook Cloud Sync (only after signing in)
After you sign in (2.8), the App synchronises your wordbook (the entries, grammar points and AI analyses you have saved, and the notes you have written) together with its review progress (mastery level, review interval, next review date, review and listening-drill counts, and each card’s rating history) tothe server of your service region (for the international region, Cloudflare D1 — see Section 5), so that signing in to the same account on another device brings back the same wordbook. Key points:
- Search history never goes to the cloud: sync covers only the wordbook and the learning data listed in this section; lookup records, search history and the AI analysis cache are outside it and stay on this device only (Section 1);
- When it runs: automatically after sign-in — after launch, a few seconds after you change the wordbook, when the App returns to the foreground, and every half hour; you can also trigger it from the account panel. Without signing in, no sync ever happens;
- Review progress is learning-behaviour data: it can reflect your study habits. It is stored on the server together with the wordbook content, used only to keep your devices in sync, and never for analysis, profiling or recommendations. The rating history records only the grade you gave at each review and the number of days since the previous one (not the time of day), so that a new device can schedule intervals from the same record. The same applies to how many cards you reviewed each day (one number per date and device, so that all your devices share one daily review allowance and one streak) and to your daily review limit setting, which sync along with the wordbook;
- Reading position: for each work in Reading, the paragraph you last reached (only the work’s identifier and the paragraph number) also syncs along with the wordbook, so you can pick up where you left off on another device. It shows which works you have read, so like review progress it is used only to keep your devices in sync, and never for analysis, profiling or recommendations;
- Transport and storage: encrypted in transit with TLS and stored on the server under the platform’s encryption at rest; the cloud copy is linked only to your account, never enters any shared cache and is never shown to other users;
- Retention: kept until you delete your account, and removed immediately on deletion (2.8). Signing out merely stops syncing and does not delete the cloud copy; signing in again resumes it;
- Portability: you can export the whole wordbook as a JSON file at any time under “Settings → Wordbook → Export”.
3. What We Do Not Do
- No required account, and no collection of identifying information such as name or telephone number; an email address is collected only when you sign in yourself (2.8);
- No embedded third-party advertising or analytics SDKs;
- No selling or renting of data, and no sharing of data with third parties for marketing;
- No reading of browsing history unrelated to lookups (the extension works only when you trigger a lookup).
4. Data Retention at a Glance
| Data | Where it is stored | Retention |
|---|---|---|
| Dictionary lookups, history, wordbook, settings | Local only (on mobile, may enter your own cloud backup via system backup, which we cannot access — see Section 1) | Controlled by you; deletable at any time |
| Anonymous device identifier and quota counters | Server | Quota counters roll over daily; the server keeps no device inventory, and the identifier appears only in logs, statistics and authorisation records, deleted with each of their retention periods |
| Speech synthesis cache (not linked to device) | Server | Long-term (content-addressed, no identity) |
| AI analysis shared cache (not linked to device) | Server | Around 90 days |
| Share authorisation records (linked to device, triggered by you) | Server | Same cycle as the shared cache; withdrawable |
| Email address (encrypted) and account record (sign-in only, see 2.8) | Server | Until you delete your account; deleted immediately on deletion |
| Wordbook cloud copy (including review progress and reading position; after sign-in, see 2.9) | Server | Until you delete your account; deleted immediately on deletion |
| User feedback | Server | At most 180 days |
| Crash and error reports | Server | At most 180 days |
| Server logs / aggregated statistics (metadata, includes country/region-level coarse region, excludes the IP address itself) | Server | Around 7 days / around 3 months |
5. Data Hosting and External Processors
We do not “share” data with third parties, but the following infrastructure and model services process data on our behalf in order to provide functionality:
| Processor | What it processes | Purpose |
|---|---|---|
| Cloudflare, Inc. | All server requests and website visits (the gateway is deployed on Cloudflare Workers, the website on Cloudflare Pages — see Section 7); storage of accounts and wordbook cloud copies (Cloudflare D1, see 2.8 and 2.9); sending of sign-in verification emails | Hosting, distribution, storage and email delivery |
| Microsoft (Azure AI Speech) | The word or sentence text being read aloud | Speech synthesis |
| OpenAI | The query text for AI analysis (on a cache miss; except requests from mainland China, Hong Kong and Macau) | Generating analysis content |
| DeepSeek | The query text for AI analysis (only requests from mainland China, Hong Kong and Macau, on a cache miss) | Generating analysis content |
| Name.com, LLC (Titan Email) | Messages you send to [email protected]; notification copies of user feedback and crash reports (2.4, 2.5) | Hosting our mailbox |
AI analysis is generated by direct connection to OpenAI’s official API; requests from mainland China, Hong Kong and Macau are instead generated by direct connection to DeepSeek’s official API, and neither serves as a fallback for the other. Text sent to the model and speech providers does not carry your device identifier; each processor handles data according to its own terms of service; the interfaces we have chosen are all intended for API use, and any default undertaking not to use data for model training is governed by each provider’s terms as they stand at the time.
6. Your Choices and Controls
- Turn off crash reports: Settings → General → Diagnostics (see 2.5);
- Do not use cloud features: if you do not tap pronunciation or AI analysis, no corresponding upload is triggered; without signing in, no wordbook sync happens; dictionary lookup, wordbook, review and other core features work fully offline;
- Sign out / delete your account: “Me / Settings → Account” (see 2.8). Signing out stops syncing and keeps your local wordbook; deletion removes your email address, account record and wordbook cloud copy immediately;
- Export your wordbook: “Settings → Wordbook → Export” writes the whole wordbook to a JSON file you can take anywhere;
- Delete local data: clear history/wordbook in the App, or uninstall it;
- Feedback and questions: contact us through the in-app feedback entry point or at [email protected].
7. Website and Dictionary Site
The sections above describe the desktop application, the mobile application and the browser extension; data handling for the website (tomoshi.app, including its download pages) and the dictionary site (dict.tomoshi.app, publicly browsable static dictionary pages) is separate and simple, as follows:
- Hosting and distribution: the website is deployed on Cloudflare Pages, and the dictionary site is served by Cloudflare Workers and R2 (both see Section 5). As with visiting any website, Cloudflare processes connection information such as your IP address when responding to requests, for content delivery and security protection; we do not otherwise record or use visitors’ IP addresses;
- Analytics: the website and the dictionary site each use a Cloudflare Web Analytics site to understand traffic. It uses no cookies, does no cross-site tracking and builds no personal profiles; when a page loads, your browser sends technical information such as the page address, referring page, and browser and device type to Cloudflare, and we see only aggregated figures (there is no analytics of any kind inside the App, see Section 3); the analytics script is not loaded for visitors whose browser time zone is in Europe (see Section 11); the dictionary site has no forms;
- Local preferences and lookup history: your interface language preference (Simplified / Traditional) is stored only locally in your browser (localStorage) and is not uploaded; the dictionary site likewise stores the words you look up on it (the query text and the page opened) only locally in your browser (at most 5,000 entries), for the “Recent” chips on its home page and the “Lookup history” panel; they are not uploaded and not synced with the Tomoshi app, and you can delete entries one by one or clear them all in “Lookup history”; neither site sets any cookie of its own.
8. Notice for Users in Japan (External Transmission Rules)
As required by the External Transmission Rules of Japan’s Telecommunications Business Act, the information transmitted from your device to external parties by the App and the website, its purpose and its recipients are itemised in Sections 2 and 7, and summarised here:
| Information transmitted | When | Recipient | Purpose |
|---|---|---|---|
| Text to be read aloud, voice | Tapping pronunciation | Our own server (synthesised via Microsoft) | Speech synthesis |
| Query text | AI analysis (including automatic triggering on a dictionary miss) | Our own server (generated via OpenAI; via DeepSeek for mainland China, Hong Kong and Macau) | Generating analysis |
| Anonymous device identifier | Each cloud request | Our own server | Quota and abuse prevention |
| Feedback content, version, system type | Submitted by you | Our own server | Handling feedback |
| Error stack, version, system type | Crash / internal error (can be turned off) | Our own server | Fixing defects |
| Update check request (no device identifier, no lookup content) | Periodically, automatically | Our own distribution server (data.tomoshi.app) | Checking for dictionary data and app updates |
| Website and dictionary site visit information (page address, referring page, browser type and so on) | Browsing a website or dictionary site page | Cloudflare (Web Analytics) | Traffic statistics |
| Email address | Tapping “Sign in with email” yourself | Our own server (verification email sent via Cloudflare) | Sign-in verification (see 2.8) |
| Wordbook, review progress and reading position | Automatically after sign-in (see 2.9) | Our own server | Cross-device sync |
9. Minors
The App is aimed at general language learners, is not designed for children, and does not knowingly collect personal information from children.
10. Changes to This Policy
When this policy is updated, we will update this page and state the effective date; substantive changes involving new data collection will be prominently notified in the App.
11. Notice for Users in the EEA, the UK and Switzerland
This section supplements the rest of this policy for users located in the European Economic Area (EEA), the United Kingdom or Switzerland, as required by the General Data Protection Regulation (GDPR), the UK GDPR and the Swiss Federal Act on Data Protection. The earlier sections apply to you as well; this section covers only what those laws additionally require us to state.
11.1 Controller and Representatives
- Data controller: Zhang Yuanyi (individual developer, based in Japan), contact [email protected].
- Representative in the EU/EEA, the UK and Switzerland (GDPR Article 27,
UK GDPR Article 27 and Article 14 of the Swiss Federal Act on Data
Protection): we have appointed Data Protection Representative Limited
(trading as DataRep) as our representative in all three. You can reach us
through DataRep in any of the following ways:
- by email to [email protected], quoting “Tomoshi / y1z” in the subject line;
- through the web form at https://www.datarep.com/data-request ;
- by post — EU/EEA: DataRep, 77 Camden Street Lower, Dublin, D02 XE80, Ireland; UK: DataRep, 107-111 Fleet Street, London, EC4A 2AB, United Kingdom; Switzerland: DataRep, Leutschenbachstrasse 95, Zurich, 8050, Switzerland. DataRep also has a postal address in every EU member state, Norway and Iceland, and you may write to the one in your own country instead (ask via the web form or email above).
- When writing by post, address your letter to “DataRep”, not “Tomoshi”, or it may not reach them, and state in the letter that it concerns Tomoshi. To protect your data, we may need to verify your identity before acting on a request.
You may contact us, or the representative, about any matter covered by this policy. For questions not about personal data (features, using your account, bug reports), please contact [email protected] directly — the representative handles personal-data requests only. Our legal representative under the EU Digital Services Act is listed in Section 15 of the Terms of Service. We have not appointed a data protection officer (DPO): the core activities of the service involve neither large-scale systematic monitoring nor special categories of data, so Article 37 of the GDPR does not require one.
11.2 Purposes and Legal Bases
| Data (see the relevant section) | Purpose | Legal basis |
|---|---|---|
| Anonymous device identifier and quota counters (2.1) | Providing the cloud features you request, daily usage quotas and abuse prevention | Performance of a contract (Art. 6(1)(b)); abuse prevention is a legitimate interest (Art. 6(1)(f): keeping the service secure and its use fair) |
| Text to be read aloud (2.2) | Speech synthesis | Performance of a contract (Art. 6(1)(b)) |
| Query text for AI analysis (2.3) | Generating the analysis | Performance of a contract (Art. 6(1)(b)) |
| Shared AI analysis cache (2.3) | Avoiding repeat generation and controlling cost | Legitimate interest (Art. 6(1)(f)); the cache is not linked to devices or accounts |
| Share authorisation records (2.3) | Crediting shares you actively make, and supporting withdrawal | Performance of a contract (Art. 6(1)(b)) |
| Email address, account record and wordbook cloud copy (2.8, 2.9) | Sign-in and cross-device sync | Performance of a contract (Art. 6(1)(b)) |
| Account usage dates (2.8) | Counting how many accounts use the service each day | Legitimate interest (Art. 6(1)(f): understanding the scale of the service; day-level only, producing an aggregate figure only) |
| User feedback (2.4) | Handling and responding to your feedback | Legitimate interest (Art. 6(1)(f)) |
| Crash and error reports (2.5) | Finding and fixing defects | Legitimate interest (Art. 6(1)(f): keeping the App stable); sent only on a crash and never containing your queries; can be turned off in Settings at any time |
| Server logs and coarse region (2.6) | Troubleshooting, capacity planning and routing to a nearby server | Legitimate interest (Art. 6(1)(f)) |
| Update check requests (2.7) | Checking for dictionary data and App updates | Legitimate interest (Art. 6(1)(f): delivering fixes and data updates promptly) |
| Website and dictionary site analytics (Section 7) | Understanding traffic | Legitimate interest (Art. 6(1)(f)); the analytics script is not loaded for visitors whose browser time zone is in Europe, so this normally does not happen for you |
You are under no statutory or contractual obligation to provide this data: if you do not use the cloud features or sign in, dictionary lookup, the wordbook and reviews work as usual; only the corresponding cloud features are unavailable. The service carries out no automated decision-making, including profiling, within the meaning of Article 22 of the GDPR.
11.3 International Transfers
We are based in Japan. The European Commission has recognised Japan as providing an adequate level of protection for personal data (2019 adequacy decision), and the United Kingdom recognises that finding as well; Switzerland has not placed Japan on its list of countries with adequate protection. For the processors listed in Section 5 that are located in other countries or regions, transfers rely on the following:
- Cloudflare, Inc. (United States, global network): certification under the EU-U.S. and Swiss-U.S. Data Privacy Frameworks (DPF), and the standard contractual clauses (SCCs) in its data processing addendum, including the UK Addendum and the provisions for Switzerland;
- Microsoft (Azure AI Speech): speech is synthesised in Azure’s Japan region; Microsoft is also certified under the EU-U.S. and Swiss-U.S. Data Privacy Frameworks, and its data processing addendum applies SCCs to transfers out of the EEA, the UK and Switzerland;
- OpenAI: under its data processing addendum, data from the EEA and Switzerland is processed by OpenAI Ireland Limited (Ireland), whose onward transfers outside Europe rely on SCCs or an adequacy decision; data from the UK is processed by OpenAI OpCo, LLC (United States) under the SCCs with the UK Addendum;
- Name.com, LLC (United States, provider of our Titan Email mailbox): the SCCs in its customer data processing addendum;
- DeepSeek (China): serves only requests from mainland China, Hong Kong and Macau (see 2.3). Requests from the EEA, the UK and Switzerland are never passed on to DeepSeek: even if OpenAI is temporarily unavailable, AI analysis for those requests simply fails rather than falling back to a provider in China. The region of a request is determined from the coarse country-level location described in 2.6; when it cannot be determined, the request likewise never falls back.
You may write to us for a copy of the relevant safeguards.
11.4 Retention
Retention periods for each kind of data are set out in Section 4.
11.5 Your Rights
To the extent provided by applicable law, you have the right to:
- access the personal data we hold about you and obtain a copy;
- rectify inaccurate data;
- erase your data — your account and wordbook cloud copy can be deleted instantly in the App (2.8);
- restrict processing;
- data portability — the wordbook can be exported to a JSON file in the App at any time (Section 6);
- object to processing based on legitimate interests — for example by turning off crash reports (2.5);
- withdraw consent — withdrawal does not affect the lawfulness of processing based on consent before it was withdrawn.
To exercise these rights, write to [email protected]. We will reply within one month of receiving your request, extendable by two further months where the law allows, in which case we will tell you why. Because most data is linked only to an anonymous device identifier, we may need information that lets us locate it (for example your sign-in email address); if we cannot match a piece of data to you, we will tell you so.
You also have the right to lodge a complaint with a data protection supervisory authority — in the EEA, the authority in the country where you habitually live or work or where the alleged infringement took place; in the UK, the Information Commissioner’s Office (ICO); in Switzerland, the Federal Data Protection and Information Commissioner (FDPIC).
The Simplified Chinese version of this policy is the original and governing version. This English version is provided for convenience only; in the event of any discrepancy between the two, the Simplified Chinese version prevails.