1. 我们收集哪些信息

• 账号信息：如邮箱地址、登录密码的加密摘要、账号角色、注册时间、最近活跃时间。
• 对话信息：你在聊天中输入的文字、系统回复、会话标题、会话时间、历史记录。
• 语音与转写信息：当你主动启用语音输入时，可能收集语音片段、转写文本与相关语音服务日志。
• 画像与分析信息：为改善服务质量，我们可能基于你的对话生成情绪状态、主题归纳、支持需求、最新能量卡、管理员分析报告等内部信息。
• 设备与访问日志：如 IP、浏览器类型、页面访问时间、错误日志、基础安全审计记录。

2. 我们为何处理这些信息

• 为你创建和维护账号、完成登录、邮箱验证和找回密码。
• 为你提供连续聊天、历史会话恢复、语音输入和能量卡展示。
• 为保障系统安全、识别滥用、排查错误与优化稳定性。
• 为改进模型效果、优化提示词、观察服务质量与产品体验。
• 为管理员在内部控制中心查看必要的运营分析、用户报告与模型优化建议。

3. 关于敏感信息与使用边界

你的聊天内容、语音输入、情绪趋势、支持需求、用户画像和分析报告，可能涉及较高敏感度的信息。我们会尽量按照最小必要原则处理这类信息，仅用于提供服务、保障安全和优化体验。普通用户端不会展示内部画像标签与管理员分析结果。

4. 数据存储地点与跨境说明

当前站点部署在中国香港。若你从中国大陆、欧洲、北美或其他地区访问本服务，你提交的账号信息、聊天信息和相关日志可能会被传输并存储至中国香港。你在使用本服务前，应确认自己已了解并接受这一点。

5. 第三方服务

为了提供核心能力，我们可能使用第三方基础设施或接口服务，例如：

• 大模型服务：用于生成聊天回复、分析报告或文本归纳。
• 语音识别服务：用于将你主动提交的语音输入转写成文字。
• 云服务器、网络安全与邮件服务：用于站点运行、发信验证和基础安全防护。

我们会尽量控制共享范围，仅在提供功能所必需时调用这些服务。

6. 我们如何保护信息

• 密码以加密摘要形式存储，不以明文保存。
• 通过管理员权限区分、会话控制和日志记录减少越权访问风险。
• 对配置、发信、用户分析等后台操作保留必要的内部记录。
• 持续改进访问控制、错误监控与安全防护措施。

7. 保存期限

在你主动删除历史对话、注销账号，或在为实现本政策目的所必需的期限届满后，我们会根据业务需要和法律义务处理相关数据。部分日志、安全审计记录和必要的配置备份可能在合理期限内继续保留。

8. 你的权利

• 你可以删除自己的历史对话。
• 你可以通过找回密码功能重置密码。
• 你可以申请注销账号或请求删除与你账号相关的信息。
• 你可以要求了解我们如何处理你的个人信息，或对明显错误的信息提出更正请求。

如需行使相关权利，请通过本网站公开的官方联系渠道与运营团队联系。

9. 未成年人

本服务当前更适合成年人使用。若你是未成年人，请在监护人知情并同意的前提下使用，并避免在没有监护人支持的情况下将本服务视为危机干预或专业帮助替代方案。

10. 政策更新

我们可能根据产品迭代、部署调整、合规要求或第三方服务变化更新本政策。更新后的版本将在本页面发布，并以更新日期作为生效参考。

11. 关于 AI 服务的特别说明（2026 年 7 月）

根据《人工智能拟人化互动服务管理暂行办法》的要求，我们在此向你说明：

• 服务性质说明：Found 提供的所有对话服务均由人工智能生成，并非真人互动。你在与 AI 对话时，请始终保持清醒认知。
• 使用时长与健康提醒：Found 倡导适度、理性使用。请你根据自身身心状态合理安排使用时长，并在感到疲劳、情绪明显波动或影响正常休息时，主动暂停使用并休息调整。我们也会结合产品能力逐步完善相关提醒机制。
• 未成年人保护：本服务不向未成年人提供虚拟亲密关系或虚拟伴侣服务。如果你未满 18 周岁，请在监护人知情同意的前提下使用。如你未满 14 周岁，需取得父母或其他监护人的同意。
• 风险提示与求助建议：Found 不是医疗服务、心理危机干预平台或紧急救助平台。当你在对话中明确表达自伤、自杀、伤人或其他紧急危险信号时，系统可能生成风险提示内容，建议你尽快联系当地紧急求助资源、可信赖的人，或有资质的专业支持机构。
• 对话数据的使用说明：你的聊天内容可能涉及敏感个人信息。除法律另有规定，或经你另行同意外，我们不会将你的聊天内容直接用于面向公众的通用模型训练。相关数据主要用于向你提供当前服务，以及在最小必要范围内用于安全保障、故障排查、服务质量评估和产品功能优化。

1. Information we collect

• Account information, such as email address, encrypted password digest, account role, sign-up time, and recent activity time.
• Conversation information, including the text you enter, system replies, conversation titles, timestamps, and history.
• Voice and transcription information when you actively use voice input, including audio clips, transcribed text, and related speech-service logs.
• Profile and analysis information, such as emotional state summaries, topic groupings, support needs, latest energy card, and internal admin reports generated from conversation history.
• Device and access logs, such as IP address, browser type, page access times, error logs, and basic security audit records.

2. Why we process this information

• To create and maintain your account, complete sign-in, email verification, and password recovery.
• To provide continuous chat, conversation restoration, voice input, and energy-card display.
• To protect service security, detect misuse, troubleshoot issues, and improve stability.
• To improve model quality, prompt strategy, service quality, and product experience.
• To allow administrators to review necessary operational analytics, user reports, and model-improvement suggestions inside the internal control center.

3. Sensitive information and boundaries

Your conversation content, voice input, emotional trends, support needs, user profile signals, and analysis reports may involve higher-sensitivity information. We aim to process such information only on a minimum-necessary basis for service delivery, security, and product improvement. Internal profile labels and admin analysis are not shown on the normal user side.

4. Storage location and cross-border notice

The current site is deployed in Hong Kong, China. If you access the service from mainland China, Europe, North America, or any other region, your account information, conversations, and related logs may be transmitted to and stored in Hong Kong, China. You should make sure you understand and accept this before using the service.

5. Third-party services

We may use third-party infrastructure or interfaces to provide core functionality, such as:

• Large-model services for replies, internal reports, and text summarization.
• Speech-recognition services for converting your voluntarily submitted voice input into text.
• Cloud hosting, security, and email services for site operation, verification emails, and baseline protection.

We aim to keep any data sharing narrow and limited to what is necessary to provide the relevant function.

6. How we protect information

• Passwords are stored as encrypted digests rather than plain text.
• Administrator permissions, session controls, and logs are used to reduce unauthorized access risk.
• Necessary internal records are kept for configuration, email, and admin analysis actions.
• We continue to improve access control, monitoring, and security protections over time.

7. Retention

We retain information for as long as needed to provide the service, support security, fulfill reasonable operational needs, or comply with applicable obligations. Some logs, audit records, and necessary backups may remain for a reasonable period even after specific content is deleted.

8. Your rights

• You can delete your own conversation history.
• You can reset your password through the password-recovery flow.
• You can request account deletion or ask that information linked to your account be removed.
• You can ask how your personal information is being handled and request correction of clearly inaccurate information.

To exercise these rights, please contact the operating team through the official contact channel published on this website.

9. Minors

This service is better suited to adults at this stage. If you are a minor, please use the service only with the knowledge and consent of a parent or guardian, and do not treat it as a substitute for crisis support or professional care.

10. Policy changes

We may update this policy to reflect product changes, deployment changes, legal requirements, or third-party service changes. The updated version will be published on this page, and the update date will serve as the reference point for effectiveness.

11. Special notice about AI services (July 2026)

In accordance with the Interim Measures for the Administration of Artificial Intelligence Anthropomorphic Interactive Services, we provide the following notice:

• Service nature: All conversation services provided by Found are generated by artificial intelligence and do not involve real human interaction. Please remain clearly aware of this when interacting with the AI.
• Usage duration and well-being reminder: Found encourages moderate and mindful use. Please arrange your usage time according to your own physical and emotional condition, and pause the service when you feel fatigued, emotionally unsettled, or notice interference with normal rest. We may continue to improve related reminder mechanisms as product capabilities evolve.
• Protection of minors: This service does not provide virtual intimate relationship or virtual companion services to minors. If you are under 18 years old, please use the service only with the knowledge and consent of your guardian. If you are under 14 years old, consent from a parent or other legal guardian is required.
• Risk notice and help-seeking guidance: Found is not a medical service, mental-health crisis intervention platform, or emergency assistance platform. When you explicitly express self-harm, suicide, violence, or other urgent danger signals in a conversation, the system may generate risk-related guidance suggesting that you contact local emergency resources, trusted people around you, or qualified professional support organizations as soon as possible.
• How conversation data is used: Your chat content may involve sensitive personal information. Unless otherwise required by law or separately consented to by you, we do not directly use your chat content for public general-purpose model training. Such data is mainly used to provide the current service to you and, within the minimum necessary scope, for security protection, troubleshooting, service-quality evaluation, and product-function improvement.