Personal Information and Emotional Data Protection Statement

Last Updated: September 12, 2026 · Effective Date: September 12, 2026 · Applicable Products: MulanXu® Full Platform · MulanXu Headquarters, Building A6, Yunhe Vanke Center, Gongshu District, Hangzhou

I. Legal Basis of This Statement

• Personal Information Protection Law of the People's Republic of China (PIPL)
• Data Security Law of the People's Republic of China
• Cybersecurity Law of the People's Republic of China
• Interim Measures for the Management of Generative Artificial Intelligence Services
• Interim Measures for the Management of AI Anthropomorphic Interaction Services (effective July 15, 2026)
• Measures for the Standard Contract for Outbound Transfer of Personal Information
• Methods for Identifying Illegal Collection and Use of Personal Information by Apps
• GB/T 35273 Information Security Technology — Personal Information Security Specification

II. Emotional Data = Sensitive Personal Information

• Emotional relationship information:intimate relationship descriptions, dating status, marital status, relationship satisfaction, and emotional tendencies toward a partner
• Chat content:conversation records, text, voice, and images (relationship-related) between users and AI or between users and real users
• Relationship preferences:values orientation, partner preferences, relationship style, and communication pattern profiles
• Psychological self-reports:user-initiated self-reported content such as depression, anxiety, stress, insomnia, family conflicts, and intimate relationship difficulties
• Behavioral trajectories:offline event participation, geographic location check-ins, and club visit records
• AI reports:AI-generated personality profiles, relationship matching suggestions, emotion recognition reports, and risk identification records
• Crisis signals:high-risk signals such as self-harm, suicide, or violent tendencies expressed by users in conversations

III. Separate Consent Mechanism

The processing of emotional data is not bundled with basic account services. If a user declines separate consent, they may still use MulanXu's basic browsing, public content viewing, and brand information features, but cannot use core services such as AI relationship companionship, AI reports, and relationship matching.

• When first enabling AI conversation:a separate pop-up informs you of the scope of emotional data, retention period, purposes, and your right to withdraw
• When first generating an AI report:a separate pop-up informs you that data is used for profile modeling and report generation, and that training use can be disabled
• When first uploading relationship-related content:a separate pop-up informs you that data is encrypted for storage and can be deleted and exported
• Processing begins only after the user clicks "Agree"; when the user clicks "Decline," no coercion is applied — only the relevant features are disabled

IV. User Rights and How to Exercise Them

• Right of access:view the categories, sources, purposes, and retention periods of collected emotional data — via "My - Data Query"
• Right of rectification:correct inaccurate emotional profiles, relationship preferences, and personal information — via "My - Profile Editing"
• Right of deletion:delete all chat records, delete individual AI reports, delete your account — via "My - Data Management"
• Right to withdraw consent:separately withdraw "consent to emotional data processing," "consent to AI training data use," and "consent to personalized recommendations" — available at any time
• Right to obtain a data copy:export all personal data (JSON / PDF format), delivered within 10 business days via email or a download link
• Right to request account cancellation:data deletion is completed within 30 days after account cancellation, except for logs required to be retained by law
• Right to complain and report:file a complaint with pr@mulanxu.com, or with the cyberspace administration or market regulation authorities

V. Data Use Rules

• Minimum necessity:only emotional data necessary to implement specific features is collected; no collection beyond scope
• Purpose limitation:emotional data is used only for relationship style profiling, AI conversation context, AI report generation, and risk identification; it is not used for secondary marketing, cross-platform profiling, or third-party advertising
• No cross-border transfer:emotional data is entirely stored within mainland China and is not transferred abroad or provided to overseas institutions
• No sale to third parties:emotional data is never sold to third parties under any circumstances
• No forced training:users may disable "Allow use for AI model training" at any time; once disabled, the data does not enter any model training process
• No forced recommendations:users may disable "Personalized recommendations" at any time; once disabled, only non-personalized content is displayed

VI. Data Retention Periods

• Chat records:retained for 12 months by default; users may delete them manually at any time, and deleted records are cleared from the production environment within 7 days
• AI reports:retained for 24 months by default and automatically deleted upon expiration; users may delete them manually in advance
• Relationship profiles:continuously updated while the account is active and deleted upon account cancellation
• Crisis intervention records:permanently retained (after de-identification) in accordance with safety and compliance requirements, and used to improve crisis identification algorithms
• Audit logs:retained for no less than 6 months in accordance with the Interim Measures for the Management of Generative AI Services
• Legal retention:data related to incidents involving litigation or regulatory investigations is retained for an extended period as required by law

VII. Special Protection for Minors

• Date of birth is verified at registration; relationship-related features are not enabled for users under 18
• If a suspected minor is identified, emotional data collection is immediately stopped and a guardianship notice is displayed
• Small amounts of inquiry-type data submitted by minors are automatically deleted within 15 days
• Parents or guardians may request the permanent deletion of all of their child's data on the Platform

VIII. Data Security Technical Measures

• Transmission encryption:TLS 1.3 end-to-end encryption
• Storage encryption:AES-256 encryption at rest, with dual-track SM4 national cryptography encryption
• Access control:role-based least privilege (RBAC) with dual-person review
• De-identification:de-identified data is used in all development, testing, and audit scenarios
• Auditing:all data access, modification, export, and deletion is fully logged and traceable to specific personnel and timestamps
• Disaster recovery:same-city active-active deployment plus remote disaster recovery, with RPO ≤ 5 minutes and RTO ≤ 30 minutes
• Third-party audits:an independent third-party security audit is conducted annually, with an audit summary published no later than the first quarter of the following year

IX. Third-Party Sharing List

The Platform does not sell emotional data to third parties and shares data only in the following minimally necessary scenarios:

• Domestic AI large models:only de-identified conversation content is transmitted for AI generation; identity information and emotional profile tags are not transmitted
• Payment channels:only order numbers and payment status are shared during payment; no emotional data is shared
• SMS/email service providers:phone numbers/email addresses are shared only when sending verification codes, notifications, and reports
• Cloud service providers:Alibaba Cloud / Tencent Cloud / Huawei Cloud nodes within China; no cross-border transfer
• In the event of a corporate merger, acquisition, or restructuring, emotional data is transferred under equivalent protection standards; if protection can no longer be maintained, users will be notified 30 days in advance and provided with an opt-out option

X. Cross-Border Data Risk Statement

XI. Updates to This Statement

• This statement may be updated in response to changes in laws and regulations and product evolution
• Material changes will be communicated 15 days in advance via in-app notifications, email, and SMS
• If a user does not agree to a change, they may export their data and delete their account before it takes effect
• Historical versions can be found in the "Document History" section of the footer

Applicable Terms for the Mulan Personality Assessment (ce.mulanxu.com)

Scope

In the context of the assessment site, this page concerns the AI-generated "Daily Mulan posts" and personality interpretation. The assessment site does not collect the "emotional relationship, intimate relationship, psychological state self-report" emotional data described for the Main Site, and does not process sensitive personal information.

Where AI Is and Is Not Used

• Where AI is used:copywriting refinement of the Daily posts and the response when you write "Today's Line."
• Where AI is not used:personality determination itself. Your five-dimension scores and 16 types are calculated by a fixed algorithm from each of your choices, do not pass through any large model, and are not changed by AI involvement.

Training Use

Your choice records and submissions are not used to train any model.

Risk Signals Such as Self-Harm

XII. Contact Us

• Data Protection Officer (DPO):pr@mulanxu.com
• User rights requests:email pr@mulanxu.com (response within 10 business days)
• Data security incident reporting:email pr@mulanxu.com (24/7)
• Mailing Address:MulanXu Headquarters, Building A6, Yunhe Vanke Center, Gongshu District, Hangzhou, Zhejiang Province
• Regulatory complaints:Illegal and Harmful Information Reporting Center of the Cyberspace Administration of China (www.12377.cn); Zhejiang Provincial Cyberspace Administration; Zhejiang Provincial Communications Administration

Special Provisions for Mulan Personality Assessment (ce.mulanxu.com)

This section applies only to the Mulan Personality Assessment (ce.mulanxu.com). Users of ai.mulanxu.com continue to follow the main body of this statement.

1. No Registration

Open it and start — no phone number, no email, no password, no real-name verification. The only things that identify you are a random anonymous ID stored in your browser (valid for one year) and an 8-digit Mulan code.

2. What We Collect

• Your choices on each question (question number, option, source) and the five-dimensional vector computed from them
• One action per day: the behavior you write down that actually happened that day (including the situation and what you did)
• Your three-level feedback on whether today's result felt accurate
• Your submissions: situations and actions you voluntarily fill in
• Optional fields: birthday (month-day), birth year, gender, birthplace, city — all may be left blank, edited, or cleared; none affect personality assessment
• Email: only saved if you actively subscribe to the daily post; double opt-in, one-click unsubscribe at any time. No email is kept if you do not subscribe.

3. What We Do Not Collect

National ID number, phone number, real name, home address, contacts, precise geographic location, device fingerprint.

4. How Location Is Handled

• We do not store IP addresses. Anti-abuse rate-limiting stores only the IP hash, which is overwritten once the time window expires.
• Weather lookup reads the country code returned by Cloudflare, used once and discarded — not written to any database.
• City is entered by you, editable at any time or cleared via a "Prefer not to say" button. Weather caching uses roughly 11-kilometer precision grids with no user identifier attached.

5. Your Submissions

Submissions go into a review pool only: they are never shown publicly to other users, never attributed to you, and never read by any page before acceptance. Please do not include real names, contact details, organization names, or any information that could identify a specific individual or organization.

6. Cookies

Only two cookies: ce_uid (anonymous identity, one year) and ce_lang (language, one year). Both are HttpOnly + Secure + SameSite=Lax and are not used for ad tracking.

7. What We Do Not Do

We do not sell your choice records; we do not provide identifiable user data to third parties; we do not build cross-platform user profiles; we do not serve personalized ads. Aggregate statistics (norms, 16-type distribution) are bucketed by score and cannot be reverse-engineered to identify individuals.

8. Age

This assessment is a general personality exploration tool with no age gate. Users under 18 should use it with parental or guardian guidance. If you visit ai.mulanxu.com to participate in the AI emotional interactive drama, the registration and age rules of that product apply.