WhizToys PLUS+ Privacy Policy
Welcome to WhizToys PLUS+ (“the App”). The App is published and operated by the Gerontechnology Research Center (“GRC”) at Yuan Ze University (YZU), Taoyuan, Taiwan. By using the App, you confirm that you have read and understood this Privacy Policy (the “Policy”). If you do not agree, please do not use the App.
1. Scope of this Privacy Policy
This Policy applies when you download, install, register for, sign in to, and use the App, and covers how GRC collects, processes, uses, retains, and protects your personal data in connection with account services, any cloud features, and related value-added services GRC offers, as actually provided. If GRC links to this page from the App or from GRC’s website, the version published here (with its “last updated” date) controls.
This Policy does not apply to third-party sites, applications, or ads operated by others, or to content you open outside GRC’s control. Their practices are governed by their own privacy policies. If you are a minor, you should use the App with a parent or legal guardian’s consent and involvement.
2. Collection, processing, and use of personal data
GRC only processes personal data in a fair manner and for defined purposes, and where GRC has a valid legal basis under applicable law (e.g. consent, contract, legitimate interests, or legal obligation, as the case may be).
(1) Categories of data GRC may collect or process
- Account and registration data — e.g. email, display name; passwords are handled by the identity service and are not stored by GRC in clear text. The App may use Amazon Web Services (AWS) Cognito or similar.
- Game and activity records — scores, ability-related metrics (e.g. reaction, concentration, muscle endurance, memory, coordination), training plans, leaderboards, and analytics, including data from WhizToys interactive mat hardware when connected.
- Health and physical measurement data — when you use the sarcopenia assessment, the Short Physical Performance Battery (SPPB), sit-to-stand tests, or similar features, GRC collects the values you enter or that the test produces, including height, weight, BMI, appendicular skeletal muscle mass index (ASMI), grip strength, muscle mass, body fat percentage, calf circumference, gait speed, and sit-to-stand counts and times, together with your history of those results. This may constitute special-category personal data under Taiwan’s Personal Data Protection Act and comparable laws. GRC collects it only when you actively use the feature and agree, and uses it to produce your own assessment results and trend charts. It is not a medical diagnosis, treatment, or prescription. GRC will not use it for research, disclose it to third parties, or combine it with marketing without your separate, explicit consent. You may request deletion of these records at any time.
- Profile and player settings — nickname, avatar, date of birth, gender, notes, and other fields you provide (as available in the product).
- Device and connection data — device type, OS, App version, logs, IP address, and technical identifiers or status for Bluetooth Low Energy and MQTT to pair and sync with peripherals.
- Camera, photo library, microphone — only when you use those features and grant OS permissions, as required for the feature, unless you separately consent to other uses. Some features send your images or audio to third-party AI services located outside Taiwan; see Section 5a, “Third-party AI services and data security”.
- Voice recordings and voice models — if you use voice interaction or personalised voice features, GRC collects your microphone recording for speech recognition. The “create a personalised voice” feature sends a recording of your own voice of roughly ten seconds or more to a third-party service in order to build a voice model (voiceprint) that can synthesise your voice. This feature is not required to use the App and runs only when you choose to enable it. You may request deletion of the voice model and the source recording at any time.
(2) Purposes of processing and use
- Provide, maintain, and improve games, training, cognitive activities, hardware connectivity, and account services; show history, charts, leaderboards, and (where applicable) cloud sync.
- Authenticate users, protect against abuse and fraud, and maintain security; comply with law and competent authorities’ requests.
GRC does not use your personal data for marketing or profiling that is clearly unrelated to these purposes, except where you have agreed or the law allows.
(3) Cookies and similar technologies
When you use web-based login or related web views in the App, GRC or its processors may use cookies or local storage for session, security, or preferences. You can adjust browser or device settings, which may limit some features.
(4) Your rights
Depending on applicable law (e.g. Taiwan’s Personal Data Protection Act, GDPR, CCPA, where applicable), you may have rights to access, rectify, delete, restrict, object, port data, and withdraw consent where processing is based on consent. In-app account deletion or export, if offered, is described in the App. You may contact GRC using the controller and contact information at the bottom of this page; GRC may need to verify your identity.
3. Protection of data
GRC uses reasonable technical and organizational measures to help protect your personal data, such as transport encryption (e.g. HTTPS/TLS, as implemented), access controls, staff training, and reviews. Please also protect your credentials, device, and second factors.
Personal data may be stored on servers operated by GRC or by cloud providers and may be located outside Taiwan (e.g. AWS regions). Where GRC transfers data across borders, GRC uses safeguards permitted by law.
GRC retains data as long as needed for the purposes above, and longer if required for legal claims, security backups, or legal retention. When retention ends, or you delete your account and no law requires further retention, GRC deletes, stops processing, or anonymizes the data as appropriate. For the main categories, GRC’s retention practice is as follows, unless the law requires otherwise:
- Account and game records — kept while the account exists; deleted from the backend within 90 days of account deletion.
- Health and physical measurement data — kept while the account exists so you can review your history; you may request deletion of this category alone at any time, and GRC will action it within 30 days.
- Recordings and images sent for recognition — processed only for as long as that request requires and not retained long term; a personalised voice model is removed when you request deletion or delete your account.
- Connection logs — kept only as needed for security and troubleshooting, as a rule no longer than 12 months.
3a. Use by minors
The App is rated 13+ on the App Store and is not directed at children under 13; GRC does not knowingly collect personal data from children under 13. A minor should obtain the consent of a parent or legal guardian before using the App or providing personal data. Where a family member, carer, or institution creates a player profile and enters health measurement data on someone else’s behalf, they must ensure that the person concerned, or their legal representative, has consented. If GRC learns that it has collected a minor’s data without the necessary consent, GRC will delete it after verification. You may also report such a case using the contact details at the bottom of this page.
4. External links
The App or web pages that GRC provides may contain links to third-party sites, partners, or ads. If you follow a link and leave the environment that GRC operates, the collection, processing, and protection of your data is governed by that third party’s policy, not this one. GRC encourages you to read their terms and privacy statements before you provide personal data.
5. Sharing with third parties
GRC does not sell your personal data. GRC may share it only as needed in these situations:
- WhizToys backend services — the App sends data needed for sign-in, game and training records, leaderboards, measurement results, device synchronization, and related features to the WhizToys backend operated by GRC or its cloud service providers. The backend processes, stores, and returns data under this Policy. Except for the AI flows described in Section 5a or a legal requirement, GRC does not send your complete account data or feature-unrelated content to third-party AI providers.
- Processors and cloud infrastructure — e.g. AWS (including Cognito), backend services, databases, caches, and messaging, only to the extent needed to run the App, under contract and legal obligations.
- Real-time / IoT — e.g. MQTT to sync with mats, devices, or services, subject to the feature implementation.
- Third-party AI services (currently in use) — OpenAI, L.L.C. is used for health analysis, training recommendations, speech recognition, handwriting and image recognition, cognitive assessment, game content, and AI voice interaction. ElevenLabs, Inc. is used for text-to-speech and custom voice features. These services may process data outside Taiwan. See Section 5a for the specific recipients, data categories, purposes, consent process, retention, and deletion information.
- Future integrations — e.g. healthcare, care, or other third-party APIs: GRC will identify the partner, data categories, and purpose when the feature is enabled, and will obtain consent or another legal basis as required.
- Legal and regulatory — GRC may disclose data where required by law, court, or a competent authority, or to protect GRC’s rights, users, or the public, within the limits of applicable law.
GRC contractually requires subprocessors to meet appropriate data protection standards. If you are concerned about a third party’s practices, also see Section 4 above.
5a. Third-party AI services and data security
Some optional features in the App use third-party AI services provided by OpenAI, L.L.C. and ElevenLabs, Inc. Before each relevant AI feature first sends personal data to either provider, the App displays a “Third-Party AI Data Use Consent” screen that identifies the recipient, the types of data to be sent, and the purpose of the transmission, and asks for your explicit permission.
You may decline or cancel. If you do not agree, the relevant data will not be transmitted and the related AI feature will not run. Other non-AI features remain available.
1. OpenAI
Health analysis, training plans, and exercise recommendations
When you request AI health analysis, a training plan, or exercise recommendations, the data flow is:
App → WhizToys backend → OpenAI → WhizToys backend → App
Depending on the feature, the backend may send the following data to OpenAI:
- Age range and gender, where required for the requested analysis.
- Body measurements and physical performance data, such as height, weight, muscle mass, grip strength, gait speed, balance, and sit-to-stand results.
- Assessment results, such as TUG, SPPB, and sarcopenia assessment data.
- De-identified game records, ability summaries, activity data, and training records.
Before transmission, the backend removes or replaces direct identifiers such as your name, email address, phone number, account identifier, authentication token, device identifier, and precise location. OpenAI does not receive these direct account identifiers through this flow.
Speech recognition
When you actively use speech recognition, the data flow is:
App → OpenAI → App
The App sends the audio recording you actively record and submit to OpenAI to:
- Convert speech to text.
- Recognise health measurement values that you speak.
- Analyse speech rate and pauses.
- Provide voice interaction features.
The audio may contain your name, age, gender, health measurements, voice characteristics, or other personal data that you choose to speak. Because OpenAI must process the recording to perform recognition, the original recording cannot be guaranteed to be fully de-identified before transmission.
Please do not include names, contact information, medical record numbers, or other sensitive information unrelated to the current feature in your recording.
Handwriting and image recognition
When you actively use handwriting or image recognition, the data flow is:
App → OpenAI → App
The App sends the handwriting screen or image that you actively draw, photograph, select, and submit to OpenAI to recognise:
- Height, weight, muscle mass, and grip strength.
- Blood pressure, blood glucose, and other health measurement values.
- Text and numbers that you actively write or photograph.
A submitted image may contain a name, face, health value, device screen, or other identifiable information. Because OpenAI must process the original image to perform recognition, the image cannot be guaranteed to be fully de-identified before transmission.
Please review the image before submission and avoid including people, documents, contact information, medical record numbers, or surroundings unrelated to the recognition purpose.
Game and cognitive AI
When you use the JOMAC cognitive assessment or AI-generated game content, the data flow is:
App → OpenAI → App
The App may send the time period, venue type, de-identified answers, results, reaction times, and the game topics or responses you actively enter or speak to OpenAI to generate questions, analyse responses, and provide interactive content.
Your name, year of birth, and personal-information questions used in JOMAC are processed on the device and are not sent to OpenAI. If you use voice interaction, the App may additionally send the audio recording, transcript, and conversation content that you actively provide.
AI voice assistant and personalised responses
When you use the AI voice assistant or game-hosting features, OpenAI may receive the voice recording, transcript, and conversation content that you actively provide to understand your request, generate a text response, and provide game interaction.
If you enable personalised responses, the App may also send your player nickname and recent game records to OpenAI to generate personalised greetings, game suggestions, and responses. Please do not provide contact information, medical record numbers, or other sensitive information unrelated to the current feature in the conversation.
If an AI-generated response is played aloud, the AI response text is also sent to ElevenLabs, Inc. for text-to-speech processing.
2. ElevenLabs
When you use text-to-speech or custom voice features, the data flow is:
App → ElevenLabs → App
Depending on the feature, the App may send:
- Text that you request to be converted into speech, including AI-generated response text.
- A voice sample that you actively record.
- A custom voice name that does not contain your real name.
- Voice and playback settings required to generate the audio.
When you create a custom voice, ElevenLabs analyses the voice sample and creates a voice model capable of reproducing its characteristics. The sample and model may constitute sensitive or biometric-related data.
Do not upload another person’s recording to create a custom voice unless you have obtained that person’s explicit permission.
3. Purposes of processing
Data sent to third-party AI services is used only to provide the feature you requested, including:
- Generating health analyses, training plans, and exercise recommendations.
- Converting speech to text and text to speech.
- Recognising handwriting, numbers, and images.
- Generating cognitive assessment questions, game content, and interactive responses.
- Creating or using a custom voice when you explicitly request that feature.
AI-generated content is provided only for health management and exercise reference and does not constitute a medical diagnosis, treatment, or professional medical advice.
4. Data security and provider protections
GRC applies or requires the following safeguards:
- Only data needed to perform the selected feature is transmitted.
- Direct identifiers are removed from structured health data where de-identification is appropriate.
- Data is transmitted using HTTPS/TLS encryption.
- Access to API keys, system accounts, and data is restricted.
- Third-party providers are bound by applicable service terms, data-processing agreements, and cross-border transfer safeguards.
- GRC periodically reviews the providers’ information-security, privacy, and retention practices.
- Data processed through AI services is not sold to third parties or used for advertising.
OpenAI states that input and output submitted through its API are not used to train or improve its models by default. Depending on the API endpoint and data-control settings used, some data may be retained for up to 30 days to prevent abuse and maintain service security, unless a longer period is required by law.
ElevenLabs may process text, voice samples, voice data, and generated audio to provide speech synthesis and custom voice services. A custom voice model must be retained while you continue to use that custom voice.
Under ElevenLabs’ published policy and depending on the plan and data-use settings of the account used by GRC, ElevenLabs may also use newly submitted text, audio, and voice data to improve its AI models if “Improve the models for everyone” has not been disabled. After it is disabled, newly submitted data is not used for that purpose.
Relevant data may be processed outside Taiwan, including in the United States or wherever a third-party provider operates its service facilities.
5. Retention and deletion
- OpenAI API data is processed according to the endpoint and data-control settings used; some abuse-monitoring logs are retained for up to 30 days by default.
- ElevenLabs data is retained according to the feature used and the applicable account settings.
- Under ElevenLabs’ published policy, voice-related data it generates is deleted no later than three years after GRC’s last interaction with the service, unless a longer period is required by law.
- Temporary recordings and images on the device are deleted after recognition is completed or cleared under the operating system’s temporary-storage rules, unless you separately save them.
- A device-side voice sample used to create a custom voice is deleted after the voice is successfully created.
- An ElevenLabs custom voice model is retained until you delete that voice or the service purpose ends.
- You can delete a created voice model through the App’s custom voice feature.
- If you withdraw AI consent, the App stops future related transmissions. Withdrawal does not affect processing lawfully completed before withdrawal.
- You may use the contact details below to request that relevant data processing stop or that relevant data be deleted.
To request that AI data processing stop or that related data be deleted, use “WhizToys PLUS+ AI Data Request” as the email subject and provide the email address needed to identify your account.
For more information, see OpenAI API data controls and the ElevenLabs Privacy Policy.
6. Location permission
The iOS version does not request location permission, and GRC does not collect location data from iOS users. Apple’s CoreBluetooth does not require location permission to scan for and connect to Bluetooth devices.
On Android, the operating system itself requires location permission in order to scan for Bluetooth Low Energy devices. The permission is used solely to discover and connect to the mat and related peripherals. GRC does not collect your location, and does not build any record of your movements, nor use location for map navigation, geofencing, or ad tracking. If you deny or turn off the permission on Android, the app may be unable to discover or connect to the mat and some features may be unavailable.
7. Changes to this Privacy Policy
GRC may update this Policy for legal, technical, or product reasons. GRC will post the new “last updated” date on this page and, where practicable, notify you through the App, in-product messages, or email. Continued use after the update may mean you accept the new Policy, as allowed by law. If you do not agree, you may stop using the App and use account-deletion or contact GRC using the information at the bottom of this page. For questions about the Policy or your data, use the same contact information.