These AI Features Terms apply to ToSche AI chat, goal consultations, image reading, voice input, in-app actions and memory provided by Synthera. You are interacting with AI, not a person. You must agree to these Terms separately from the general ToSche Terms. If you decline or withdraw agreement, you cannot use AI, but non-AI features remain available under their normal conditions. The general ToSche Terms govern fees, cancellation, liability, changes and dispute resolution not addressed here. Section numbers below are shared with the general ToSche Terms.
AI features are available from age 13. If you are under 18, you may not use AI features without the consent of a parent or other legal guardian. You also need their consent to purchase a Paid Plan. State your age accurately. For ages 13 to 17, we apply age-appropriate safety instructions and do not save especially sensitive information, such as health information, to AI memory.
1. To use the AI Features (the AI assistant and goal-setting chats within it, including goal coaching in versions earlier than 2.9.0) and voice input, you need to review and agree to the consent screen we show before data is first sent to the AI Features. Without consent, you cannot use the AI Features or voice input. You may withdraw consent at any time from "Settings" → "AI & Privacy"; after withdrawing, you cannot use the AI Features or voice input until you consent again. 2. When you use the AI Features, you agree that the following data is sent to Anthropic, PBC (a Delaware corporation, United States of America), the provider of the AI models: • Text you enter in the AI chat • Related information the AI reads in order to answer or carry out an operation, such as tasks, schedules (including calendar events and events imported from the iPhone Calendar), and goals (titles, dates/times, locations, notes, contents, etc.) • The names of calendars and labels • AI memory (including sensitive personal information saved with your consent) and the AI shared memo • In goal-setting chats (version 2.9.0 and later), the goal being discussed and its steps, the tasks linked to the goal, and summaries of earlier goal-setting chats (an excerpt of the last reply and a record of the actions the AI took) • When AI memory is organized automatically, part of your message and the beginning of the previous reply (this feature is being rolled out gradually), and trends in how you use the App (such as the usual color for each event title) • If you are aged 13 to 17, that fact (sent together with the safety instructions in Section 2-2, paragraph 4) • Recent chat history to keep the context of the conversation • Your device's time zone and the App's language setting • Search terms, if the AI uses web search (sent through Anthropic to the search service it uses) 3. When you use voice input, you agree that the audio you record is sent to Google LLC (United States of America), the provider of the speech recognition service. The audio is used only to convert it to text and is not stored on our servers. 4. The details of the transmissions in the preceding 2 paragraphs and each recipient's privacy policy are as set out in the Privacy Policy (Sections 4-2 and 4-3). These recipients are businesses in a foreign country (the United States of America), and the information we show when asking for your consent (the recipients' country, its system, and the measures the recipients take) is set out in Section 4-5 of the Privacy Policy. Anthropic, PBC has stated publicly that it does not use data received via its API to train its AI models. 5. The AI Features save sensitive personal information to AI memory only if you clearly ask and tap the consent button shown in the App. Memory saved this way is sent to the recipient in paragraph 2 each time you later use the AI Features. You can delete it at any time from "What AI remembers" in the App ("Reset AI memory" in versions earlier than 2.9.0). However, sensitive personal information is not saved to AI memory for people aged 13 to 17 (Section 2-2) or for people whose store country is the United States. Details of how AI memory is handled are set out in Section 4-2 of the Privacy Policy.
1. AI Output is provided as reference information. We do not guarantee that AI Output is accurate, complete, useful, fit for a particular purpose, up to date, or lawful. 2. AI Output may contain incorrect information, inappropriate suggestions, or content that is not factual (so-called hallucinations). Check the content of AI Output yourself, and verify it with other sources where needed, before using it. 3. AI Output is not a substitute for professional judgment in fields such as medicine, law, tax, finance, or psychology. For these, please consult a qualified professional. 4. Our liability for damage to you caused by AI Output, or by the AI Features adding, changing, or deleting events or tasks based on AI Output, is as set out in Section 14. 5. If you believe that AI Output contains harmful, illegal, discriminatory, sexually inappropriate or harassing content, improper disclosure of personal information, or harmful misinformation, you may report it to us from that AI Output's menu ("More" → "Report this response"). When you report, the category you selected, anything you chose to write, and the text of that AI Output are sent to us. We review these to check safety and act where needed, and we do not use them to train AI models. Because a report is kept as the record we act on, you cannot withdraw it yourself (if you delete your Account, the report is deleted in accordance with Section 15). We are not obliged to tell you the outcome of each report, but we will take the measures required under these Terms and Section 15-4 in respect of the reported AI Output. 6. Facts the AI Features learn from conversations and save (AI memory) and information the AI looks up on the web and shows you (such as class period times or the last day of a term) are also AI Output and may contain errors. The AI Features may add events or deadline tasks based on AI memory or information they looked up. Please check the notice shown when the AI updates its memory and the dates shown before repeating deadline tasks are added; if anything is wrong, you can correct or delete memory in "What AI remembers" in the App, and events and deadline tasks on their own screens.
When using the AI Features, in addition to Section 10, you must not do any of the following. 1. Generating, requesting, or spreading content that is illegal, harmful, threatening, abusive, harassing, defamatory, or contrary to public order and morals 2. Generating or requesting child sexual abuse material or content that sexualizes minors 3. Generating, requesting, or spreading content that encourages suicide, self-harm, violence, or harm to others 4. Unlawfully collecting, identifying, or disclosing personal information 5. Creating or spreading fraud, phishing, impersonation, or false information 6. Creating or providing malware, viruses, or other harmful code 7. Using AI Output to replace professional judgment in fields such as medicine, law, tax, or finance without consulting a qualified professional 8. Automatically sending large numbers of requests, improperly consuming AI Credits, or intentionally exhausting our server resources 9. Reverse-engineering the AI Features, prompt injection, or otherwise intentionally exploiting weaknesses in the system 10. Uses that violate the terms or usage policies (Acceptable Use Policy) of AI model providers (such as Anthropic, PBC) 11. Uses similar to the above that interfere with the safe provision of the AI Features If there is a use described above, we may restrict your use of the AI Features or take measures under Section 11. We are not responsible for damage to you caused by such measures taken for reasons attributable to you.
AI can create and edit events, assignments, tasks and goals within your request and chosen approval mode. You can choose to review changes or delegate ordinary private additions and edits. Deletion, sharing and certain other actions still require confirmation. AI can misread or make mistakes, so review saved results. Grants, consumption, expiry and transitional protections follow Section 5 and the supplementary provisions of the ToSche Terms and the subscription information. Agreeing here does not change your existing balance or expiry.
You can withdraw agreement in Settings > AI & Privacy. New AI input and voice input stop after withdrawal. Deleting conversations or memory, revoking connected services and canceling subscriptions require their respective actions. Contact support@synthera.jp.