Trust Score & Risk Scoring Policy

1. Purpose & Strategic Role

This policy defines the framework for evaluating user trustworthiness and risk levels within Glancers.

The Trust Score system is a risk-and-reputation scoring mechanism that:

• Enhances platform security

• Improves decision-making (matching, payouts, visibility)

• Rewards reliable behavior

• Detects and mitigates fraud risks

It transforms user activity into actionable trust signals across the platform.

2. Trust Score Model

2.1 Definition of Trust Score

The Trust Score is a dynamic numerical indicator representing a user’s overall reliability, based on behavioral, financial, and professional performance.

The score may be updated as new activity is recorded.

2.2 Multi-Dimensional Scoring Framework

The Trust Score is calculated across multiple dimensions:

Behavioral Dimension: - Response time - Deadline adherence - Policy compliance

Financial Dimension: - Payment success rate - Dispute frequency - Transaction consistency

Professional Dimension: - Client ratings - Project completion volume - Skill specialization consistency

2.3 Data Weighting & Recency

• Recent activity is weighted more heavily than historical performance

• High-value transactions carry greater influence

• Sustained consistency is rewarded over short-term spikes

This ensures the score reflects current reliability, not legacy performance.

2.4 Algorithm Confidentiality

• The detailed scoring logic, weights, and thresholds are strictly confidential

• Disclosure is limited to high-level guidance only

To balance transparency and integrity, Glancers provides:

• Proactive guidance signals to users (e.g., response delays, dispute frequency)

• Actionable recommendations to improve Trust Score without exposing core algorithms

This prevents:

• System manipulation

• Reverse engineering of ranking signals

3. Risk Scoring & Detection

3.1 High-Risk Signals

The system detects and reacts to suspicious behaviors, including:

• Sudden geographic changes

• Repeated failed payment attempts

• Rapid dispute escalation patterns

• Abnormal activity spikes

Context-aware safeguards are applied to reduce false positives, including:

• Differentiation between legitimate travel (e.g., digital nomads) and account compromise

• Device consistency and behavioral continuity checks

Such signals may trigger:

• Trust score reduction

• Temporary restrictions

3.2 Network Risk Analysis

User risk is influenced by connected accounts through:

• Shared devices or networks

• Transaction relationships

• Behavioral similarity patterns

An association with a high-risk account is not, by itself, a basis for reducing a User's Trust Score; network signals may prompt a review, and any score impact requires that review to find the User's own involvement.

3.3 Identity Verification Impact (KYC/KYB)

• Verified users receive higher baseline trust levels

• Additional verification increases system confidence

Unverified or partially verified accounts may:

• Face limitations

• Be subject to stricter monitoring

3.4 Behavioral Anomaly Detection

Machine learning models monitor:

• Deviations from normal usage patterns

• Non-human interaction signals

• Automated or scripted behaviors

Detected anomalies may result in:

• Score penalties

• Security reviews

3.5 Human Review & Appeal Mechanism

In cases where Trust Score impacts critical user rights (e.g., account restriction, restriction affecting compensation processing or payout processing):

• Users may request human review

• Dedicated risk teams will evaluate edge cases

• Final decisions may override automated outputs when justified

This ensures fairness and safeguards against AI misclassification

4. Trust Score Impact on Platform Privileges

Trust Score directly influences user capabilities, including:

• Visibility in search and ranking

• Access to high-value projects

• Payout-processing speed, review requirements, and applicable operational limits

• Bidding capacity

• Eligibility for premium features

Higher scores unlock enhanced privileges.

Lower scores may result in restrictions.

4.1 Trust Recovery & Rehabilitation

Glancers supports structured recovery of Trust Score through:

• Consistent positive behavior over time

• Reduced dispute activity

• Improved responsiveness and compliance

The system applies:

• Gradual score restoration mechanisms

• Decay of past negative signals when sustained improvement is detected

This ensures users are not permanently penalized for recoverable behavior

5. Fairness, Bias & Safeguards

Glancers is committed to ensuring scoring fairness:

• Periodic review of scoring models, where implemented

• Monitoring for geographic or linguistic bias

• Prevention of unfair penalization

The system is designed to reflect behavior, not identity.

6. Governance Principle

Trust Score is treated as a quantifiable, behavior-based metric within Glancers, not a financial asset or a form of property belonging to the User.

• Earned through consistent behavior

• Lost through risk signals and violations

• Continuously recalculated

The Trust Score system ensures that:

• Honest users are rewarded

• Risky behavior is contained

• Platform integrity is preserved

Procedural Review & Appeals

Any adverse administrative action, temporary restriction, or account decision taken under this policy is subject to procedural review and may be appealed in accordance with the Appeals & Reconsideration Policy.

Version History

• 2026-09 — Qualified descriptions of Platform features and service targets. Clarified enforcement, temporary restrictions, and appeal rights. Editorial and formatting corrections.
• 2026-09 — Revised selected provisions for accuracy and consistency with related policies.
• 2026-09 — Clarified consistency with related policies and qualified statements about Platform features.
• 2026-09 — Updated payment, compensation, and relationship terminology to reflect that Clients purchase Services from Glancers and that Glancers separately engages and compensates Freelancers.