1. Assessment Disclaimer
The SymbolMAG personality assessment provided through Zelfium is designed for personal insight and self-exploration. It is important that you understand the following before interpreting your results:
- The SymbolMAG personality assessment is not a clinical or diagnostic tool. It has not been designed, validated, or intended for use in clinical diagnosis of any psychological condition or disorder.
- Assessment results are not a substitute for professional psychological evaluation, therapy, or counseling. If you are experiencing psychological distress or have concerns about your mental health, please consult a qualified mental health professional.
- Personality is complex and multifaceted. No single assessment, including SymbolMAG, can capture the full scope of an individual's personality, character, or potential.
- Results represent probabilistic tendencies and preferences, not deterministic labels or fixed traits. Your results describe patterns and inclinations, not absolute categories.
- Individual results may vary across test administrations and may be influenced by life circumstances, mood, environment, and personal growth over time.
- The assessment is intended as a tool for reflection and self-awareness. It should not be used as the sole basis for major life decisions, including career changes, relationship decisions, or educational choices.
2. AI-Generated Content Disclaimer
Zelfium uses artificial intelligence to power several features across the platform. Please be aware of the following regarding AI-generated content:
- AI features within Zelfium, including Zelfium Affinia and Zelfium Libra, are powered by OpenAI GPT-4o.
- Career recommendations generated by Zelfium Libra are suggestions based on personality dimensions, not predictions of success. Many factors beyond personality influence career outcomes, including skills, experience, market conditions, and personal circumstances.
- Compatibility scores generated by Zelfium Affinia are approximate indicators, not definitive measures of relationship potential. Human relationships are far more nuanced than any algorithmic score can capture.
- AI may produce outputs that are inaccurate, incomplete, outdated, or that reflect biases present in its training data. AI-generated content should be evaluated critically and not taken at face value.
- AI-generated content does not constitute professional career counseling, relationship therapy, or psychological advice. For professional guidance, please consult qualified practitioners in the relevant field.
- Users bear sole responsibility for any decisions made based on AI-generated output. Zelfium provides these features as informational tools, not as authoritative guidance.
- AI outputs do not necessarily reflect the views, opinions, or endorsements of Zelfium Inc. The AI generates responses autonomously based on its training and the input it receives.
3. Data Handling for AI Features
Transparency about how your data interacts with AI systems is fundamental to our approach. Here is what you should know:
- What data is sent to OpenAI: When you use AI-powered features, your personality scale scores (aggregated dimensional scores, not individual item responses) and your chat messages are transmitted to OpenAI for processing.
- Zero-retention API policy: OpenAI operates under a zero-retention API policy for Zelfium's integration. This means that inputs sent to OpenAI through our API are not stored by OpenAI and are not used for model training or improvement.
- Raw responses are never shared: Your individual, raw assessment responses (your specific answers to each question) are never sent to OpenAI or any external service. Only aggregated scale scores are used.
- For complete details on how we collect, store, process, and protect your data, please refer to our Privacy Policy.
4. Regulatory Compliance
Zelfium is committed to responsible and transparent AI practices in accordance with emerging regulatory frameworks worldwide.
EU AI Act — Risk Classification
Zelfium discloses AI involvement transparently throughout the platform. Users are always clearly informed when they are interacting with AI-generated content or AI-powered features. The following risk classifications apply under the EU AI Act:
- Zelfium Libra (Career Pathing): Potentially HIGH-RISK under Annex III, Category 4 ("AI systems intended to be used for making decisions affecting terms of employment-related relationships") when deployed by organizations for employment-related decisions. When used by individual consumers for personal career exploration, Libra functions as an informational tool. Organizations deploying Libra must comply with the obligations set forth in our B2B Terms, including the Deployment Purpose Declaration.
- Zelfium Affinia (Relationship Compatibility): Classified as a limited-risk AI system subject to transparency obligations under Article 50 of the EU AI Act. Users are informed that compatibility scores are AI-generated.
- SymbolMAG Assessment: Involves automated profiling under GDPR Article 22. Users are entitled to human intervention and may contest automated profiles. See our Privacy Policy for details on your rights regarding automated decision-making.
Our Acceptable Use Policy explicitly prohibits uses that would elevate the risk classification of our tools, such as using assessment results for employment or insurance decisions.
Special Category Data (GDPR Art. 9)
Personality assessment results generated by SymbolMAG may constitute special category data under GDPR Article 9, as they may reveal psychological characteristics. Zelfium processes this data exclusively on the basis of your explicit consent (Art. 9(2)(a)) and applies enhanced safeguards including encryption, access controls, and regular data protection impact assessments. For full details, see our Privacy Policy.
Intellectual Property & Methodology Transparency
Zelfium is committed to transparency about its use of AI and automated profiling. However, the specific methodology underlying the SymbolMAG assessment — including scoring algorithms, dimensional weights, and normative data — is proprietary and protected as a trade secret. Only general descriptions of the assessment methodology are made publicly available. This balance ensures that users understand how their data is processed while protecting the intellectual property that enables the assessment's reliability and validity.
Global AI Transparency Standards
- Zelfium proactively complies with emerging AI transparency standards across jurisdictions, including clear disclosure of AI involvement, meaningful human oversight, and user control over AI interactions.
- We continuously monitor regulatory developments in AI governance and update our practices accordingly.
5. Contact
We welcome your questions and feedback regarding our assessment methodology, AI features, and the information provided in this disclaimer.
- Assessment methodology and AI features: info@zelfium.com
- Privacy and data handling concerns: contact@zelfium.com
- Company: Zelfium Inc., 1-22-11 Ginza, Chuo-ku, Tokyo, Japan
Please also review our Privacy Policy, Terms of Service, Acceptable Use Policy, B2B Terms, and Data Processing Agreement for additional terms and information governing your use of Zelfium.
Effective Date: March 31, 2026