How We Use Your Data to Improve Adjudica.AI
At Glass Box Solutions, transparency is core to our mission. This notice explains how we use information from your interactions with Adjudica.AI to improve our product—and, just as importantly, what we never use.
What We NEVER Use for Product Improvement
Protected Health Information (PHI)
- Medical records and patient information
- Any content that could identify a specific patient
- Health conditions, diagnoses, or treatments
Confidential Legal Information
- Your case strategies and work product
- Client-specific legal analysis
- Settlement positions or negotiations
- Any information protected by attorney-client privilege
These categories are used ONLY for the specific legal matter you're working on. Period.
What We MAY Use to Improve the Product
Instructional Queries & Methodological Questions
When you ask questions about how to analyze something—without case-specific details—we may use those patterns to improve Adjudica.AI.
Example of what we may learn from:
"What's the difference between the DRE method and ROM method for assigning WPI to lumbar spine injuries?"
This type of methodological question helps us:
- Improve our documentation and help content
- Enhance AI explanations of complex topics
- Identify areas where users need more guidance
Example of what we will NOT use:
"In the Martinez case, the QME assigned 12% WPI using DRE—is that correct given the L4-L5 herniation?"
This contains case-specific information and is treated as confidential.
Usage Patterns (De-Identified)
We track how features are used across all users (not tied to you specifically):
- Which tools are most popular
- Where users encounter errors
- How long analyses typically take
- Common workflow patterns
- Feature adoption rates
- Session duration and frequency
- Navigation paths through the application
- Error rates and types
What "de-identified" means:
- No user names or firm identifiers
- No IP addresses or location data
- No case-specific content
- Aggregated across all users (not individual tracking)
- Cannot be reasonably linked back to you or your firm
This helps us:
- Fix bugs and improve reliability
- Prioritize new features
- Optimize the user interface
- Improve application performance
- Enhance accessibility and usability
Specific Analytics Categories
Performance Metrics
- Page load times
- Search query response times
- Document processing speeds
- System errors and failures
CCPA Category: Internet or other electronic network activity information
Feature Engagement
- Tool usage frequency (which tools, how often)
- Feature discovery patterns
- Help documentation access
- Settings configurations
CCPA Category: Internet or other electronic network activity information
Error & Bug Reports
- Error messages and stack traces (sanitized)
- Browser and device information
- Steps leading to errors (without case content)
- Crash reports
CCPA Category: Internet or other electronic network activity information
General Feedback
When you submit feedback about the product (not about specific cases), we may use it to:
- Address common pain points
- Improve features you've commented on
- Develop new capabilities
- Inform product roadmap decisions
- Enhance training materials
What we look for in feedback:
- Feature requests
- UI/UX improvement suggestions
- Workflow efficiency ideas
- Documentation clarity issues
What we exclude from product improvement:
- Case-specific complaints or issues
- Client names or identifiers
- Confidential legal information
AI Training: Our Commitments
We Do NOT Train AI Models On:
- Your medical records or PHI
- Your case files or legal work product
- Any client-identifiable information
- Your firm's specific usage patterns
Our AI Provider (Google):
- Is contractually prohibited from using your data for training
- Does not retain PHI beyond processing your request
- Is bound by a Business Associate Agreement
We MAY Use for AI Improvement:
- General legal knowledge from public sources
- Aggregated error patterns (to improve accuracy)
- Methodological questions (without case context)
Data Retention for Analytics
| Data Type | Retention Period | Reason |
|---|
| Performance metrics | 24 months | Trend analysis and optimization |
| De-identified usage patterns | 36 months | Product improvement and research |
| Aggregated error logs | 12 months | Bug tracking and resolution |
| General feedback | Indefinitely | Product roadmap and feature development |
| Methodological questions (opted-in) | 36 months | Documentation improvement |
After retention period: Data is permanently deleted or further aggregated to remove all potential identifiers.
Third-Party Analytics Tools
We use industry-standard analytics services to collect and analyze product usage:
| Tool | Purpose | Data Shared |
|---|
| [Analytics Provider] | Usage tracking | De-identified usage patterns only |
| [Error Tracking] | Bug monitoring | Sanitized error logs (no PHI/case data) |
| [Performance Monitoring] | Speed optimization | Performance metrics only |
All third-party providers:
- Are bound by written contracts prohibiting use of data for their own purposes
- Are prohibited from receiving PHI or case-specific information
- Must comply with CCPA service provider requirements
- Are regularly audited for compliance
Your Choices
Opt Out of Product Analytics
You can opt out of having your instructional queries used for product improvement:
How to opt out:
- Email: privacy@adjudica.ai
- Subject: "Opt-Out of Product Analytics"
- Include: Your account email and firm name
- We'll confirm within 5 business days
Or use in-app controls:
- Settings → Privacy → Product Analytics → Toggle "OFF"
Note: This does not affect:
- PHI protections (which are absolute and not optional)
- Basic operational data needed to run the service
- Security and audit logging (required for compliance)
- De-identified, aggregated analytics (cannot be traced to you)
What Opting Out Means
If you opt out:
- ✓ Your methodological questions won't be used to improve documentation
- ✓ Your feedback won't inform product development
- ✓ Your individual usage patterns won't be analyzed
- ✓ You'll still receive the same service quality
- ✓ PHI and legal confidentiality remain protected regardless
What continues (cannot be opted out):
- Aggregated analytics (across all users)
- Security logging (required by law)
- Billing and account information
- Audit trails for compliance
Your California Privacy Rights (CCPA/CPRA)
As a California business using Adjudica.AI, you have specific rights under CCPA:
Right to Know
You can request:
- What categories of analytics data we collect about your usage
- How we use that data for product improvement
- Whether we "sell" or "share" your data (we do not)
Right to Delete
You can request deletion of:
- Your account and associated usage data
- Historical analytics tied to your account
- Methodological questions you've submitted
Exceptions: We may retain data required by law or for legal compliance.
Right to Correct
You can request correction of:
- Inaccurate account information
- Incorrectly attributed feedback
Right to Opt-Out
You can opt out of:
- Use of your queries for product improvement
- Non-essential analytics tracking
To exercise these rights: Email privacy@adjudica.ai with your request.
Response time: We'll respond within 45 days (extendable to 90 days if complex).
The Bottom Line
| Data Type | Used for Product Improvement? |
|---|
| Medical records / PHI | NEVER |
| Case strategies / Work product | NEVER |
| Client-specific analysis | NEVER |
| General methodology questions | Yes (without case context) |
| Feature usage patterns | Yes (de-identified) |
| Product feedback | Yes (non-case-specific) |
Compliance & Legal Basis
HIPAA Compliance
Analytics that involve PHI are governed by HIPAA Privacy and Security Rules:
- No PHI in analytics: We never include PHI in product improvement analytics
- De-identification standard: We follow HIPAA's de-identification requirements (45 CFR 164.514)
- Business Associate obligations: Our AI providers cannot use PHI for training
CCPA/CPRA Compliance
Analytics data is treated as follows under California law:
- Service provider restriction: Third-party analytics tools are service providers and cannot use data for their own purposes
- No sale or sharing: We do not sell or share your data
- Consumer rights: You can exercise all CCPA rights (know, delete, correct, opt-out)
California State Bar Ethics
For attorney users, we ensure compliance with:
- Rule 1.1 (Competence): Analytics help us improve accuracy and reliability
- Rule 1.6 (Confidentiality): No confidential client information in analytics
- November 2023 AI Guidance: We don't train on your work product
Changes to This Notice
We may update this Analytics Notice from time to time. Changes will be:
- Posted to this page with new effective date
- Emailed to active users (for material changes)
- Announced in-app (for significant changes)
Last Updated: [INSERT DATE]
Version: 1.0
Related Policies
For comprehensive information about data practices:
- Privacy Notice: Complete data handling practices
- Security Policy: How we protect your information
- Terms of Service: Legal agreement for use
- Business Associate Agreement: HIPAA compliance (for covered entities)
Questions?
Privacy Team: privacy@adjudica.ai
Privacy Officer: Alex@Adjudica.ai
Data Protection Questions: dpo@glassboxsolutions.com
We're committed to being transparent about how we use data. If you have questions about any specific type of information, please ask.
Response guarantee: We'll respond to privacy inquiries within 5 business days.
Document Information
Document Type: Product Analytics & Improvement Notice
Audience: Adjudica.AI Users (California Workers' Compensation Attorneys)
Scope: Product analytics and non-case-specific data only
Related Regulations: HIPAA, CCPA/CPRA, California State Bar Rules
Key Commitments:
- ✓ No PHI or case data in product analytics
- ✓ No AI training on confidential legal information
- ✓ Transparent about what we collect and why
- ✓ Easy opt-out for non-essential analytics
- ✓ Respect for attorney-client privilege
This Product Analytics Notice is part of our commitment to transparency. For complete details, see our Privacy Notice and Data Usage Policy.
Glass Box Solutions, Inc.
@Developed & Documented by Glass Box Solutions, Inc. using human ingenuity and modern technology