Qloo AI Governance Policy
Last Updated: 1.27.2025

Qloo Inc. ("Qloo," "we," "us," or "our") is committed to responsible AI development and deployment. This AI Governance Policy outlines our approach to AI ethics, transparency, model training, compliance, and accountability to ensure that our AI systems align with best practices while maintaining business advantages.

1. Purpose & Scope

This policy applies to all AI models, machine learning algorithms, and automated decision-making systems used within Qloo's services, including TasteDive. It governs how we collect, process, and use metadata, sentiment classifications, and engagement signals for AI training and commercial applications. It also sets the foundation for responsible AI deployment and continuous improvement.

2. Core Principles of AI Governance

Our AI governance is guided by the following principles:

• Identity Agnostic AI: Our AI models are strictly agnostic to user identity and do not incorporate or process personally identifiable information ("PII"). Identity is never used for personalization, targeting, profiling, or monetization.
• Cultural Comprehension & Entity Co-Occurrence: Our AI is designed solely to deepen understanding of cultural preferences and the relationships between entities such as movies, books, music, restaurants, and brands.
• Data Integrity & Transparency: We ensure that AI models are trained exclusively on anonymized, structured metadata, with clear documentation on data sources and processing methodologies.
• Ethical & Fair AI: We employ rigorous model validation to mitigate bias, ensure cultural neutrality, and maintain AI outputs that respect diverse global perspectives.
• Security & Compliance: Our AI governance framework adheres to industry security standards and data protection laws, including GDPR, CCPA, and emerging AI regulations.

3. Data Used for AI Training

Qloo’s AI models are trained on a diverse set of non-PII data sources, including:

• User-Generated Metadata: Structured contributions such as entity classifications (e.g., a restaurant’s "ambiance" or a brand’s "luxury" status).
• Entity Contributions: Users may submit new entities (e.g., hotels, restaurants, music artists) along with descriptive metadata.
• Sentiment & Engagement Data: Explicit sentiment inputs (positive, neutral, negative) and behavioral engagement actions (e.g., "saved" or "favorited" entities).
• Aggregated Relationship Data: Co-occurrence patterns between entities to improve AI-driven recommendations.
• Third-Party Enrichment Data: Curated, non-personal datasets used to enhance AI-driven insights.

No AI model processes, stores, or analyzes emails, usernames, IP addresses, or any form of user identity data.

4. AI Transparency & Explainability

We are committed to making our AI outputs transparent and interpretable by:

• Providing explanatory insights into how recommendations are generated.
• Maintaining model documentation that outlines data sources, methodologies, and validation techniques.
• Ensuring AI-generated results can be contextually explained without exposing proprietary model logic.

5. Bias Mitigation & Fairness

We take a proactive approach to ensure AI fairness by:

• Regularly auditing AI models to detect unintended biases in recommendations.
• Applying diverse training data to prevent overrepresentation of specific perspectives.
• Using adversarial oracles (e.g., public cultural trends, aggregate user feedback) to refine model outputs.
• Providing controls for users to refine AI-generated recommendations based on personal preferences.

6. AI & Commercial Use Cases

Qloo reserves the right to commercially leverage AI-generated insights, including but not limited to:

• Personalization Engines: Enhancing recommendations for consumers and enterprise clients.
• Predictive Analytics: Identifying cultural trends, audience behaviors, and affinity scores.
• Advertising & Brand Relevance: Generating anonymized insights for brands and advertisers.
• Enterprise Licensing: Offering structured AI-driven insights as a commercial service.

All AI-derived outputs used for commercialization are anonymized and aggregated to ensure compliance with data privacy regulations.

7. Security & Compliance

To ensure responsible AI deployment, we implement:

• Data Encryption: AI training datasets are secured through encryption protocols.
• Access Controls: Strict internal policies governing who can access and modify AI models.
• Regular AI Audits: Compliance evaluations to detect risks, biases, and security vulnerabilities.

Qloo’s AI models are designed to comply with all relevant data protection laws, including GDPR, CCPA, and emerging AI-specific regulatory frameworks.

8. Governance & Oversight

Qloo maintains an internal AI Governance Committee responsible for:

• Evaluating AI risks, compliance, and ethical considerations.
• Conducting periodic audits of AI model integrity and fairness.
• Ensuring AI governance policies remain aligned with regulatory developments.

Qloo maintains structured AI documentation covering:

• AI Model Training Data Sources: Ensuring transparency on non-PII datasets used.
• Compliance Alignment: Demonstrating adherence to legal and ethical AI principles.
• Intellectual Property Protections: Defining AI ownership and proprietary methodologies.
• Data Licensing & Commercial Agreements: Documenting AI-generated insights that are monetized through enterprise partnerships.

9. Updates to This AI Governance Policy

We may update this AI Governance Policy periodically to reflect regulatory changes, technological advancements, or internal AI policy enhancements. Continued use of the Service constitutes acceptance of any updates.

10. Contact Information

For inquiries related to AI governance or compliance, contact: [email protected]