(“Trellis Intelligence Terms”)

Effective Date: Feb 12th 2026

These Trellis Intelligence Terms (“AI Terms”) govern your use of the Trellis Health artificial intelligence features, including conversational guidance, health summaries, document interpretation, and pattern recognition tools (collectively, the “Trellis Intelligence” or “AI Services”). These terms are intentionally written in plain language to communicate transparently, if you have any questions please contact support@jointrellishealth.com.

By using Trellis Intelligence, you agree to these Trellis Intelligence Terms in addition to the Trellis Health General Terms of Service, Privacy Policy and Trellis Intelligence Privacy Policy.

If you do not agree, do not use the AI features.

1. The Role of Trellis Intelligence

Trellis Intelligence is a health information and care-support system designed to help you:

• understand medical records and lab reports
• organize your health information
• prepare for care
• recognize potential health concerns
• navigate the healthcare system
• communicate with clinicians

Trellis Intelligence provides educational and contextual health information.

It does NOT provide medical care.

Trellis Intelligence:

• does not diagnose disease
• does not prescribe treatment
• does not provide medication instructions
• does not replace a clinician’s medical judgment

Only licensed clinicians can provide medical care.

2. Relationship to Clinicians

Trellis Health provides access to clinical professionals separately from the AI.

The AI may recommend connecting you to a clinician.

A clinician will only see your conversation if you explicitly authorize it.

A clinician-patient relationship is created when you interact directly with a licensed provider — not when you interact with the AI.

3. Not for Emergencies

Trellis Intelligence is not an emergency service. Do not rely on it for urgent or life-threatening conditions.

Call 911 or seek immediate medical attention for:

• chest pain
• difficulty breathing
• stroke symptoms
• severe bleeding
• late-pregnancy bleeding or severe abdominal pain
• suicidal thoughts
• infant emergencies
• loss of consciousness

The AI may encourage you to seek care, but it cannot monitor you or dispatch help.

4. How the AI Works

Trellis Intelligence analyzes:

• your messages
• records you upload
• lab results
• structured health data

It uses machine-learning systems to generate responses.

Important limitation:

AI systems can make mistakes.

Outputs may:

• be incomplete
• contain inaccuracies
• lack full clinical context
• not reflect your entire medical history

You agree not to make medical decisions based solely on AI responses.

5. Safety Escalation

Trellis Intelligence is designed to identify potential risk situations.

If concerning patterns are detected, the system may:

• recommend clinician involvement
• offer to connect you to care
• flag for outreach

However: Trellis Intelligence cannot guarantee detection of risk or medical problems. You remain responsible for seeking appropriate medical care.

6. Appropriate Use

You may use Trellis Intelligence only for your own personal health management.

You may not:

• use it to make medical decisions for other people
• use it as a substitute for professional medical judgment
• attempt to generate prescription instructions
• rely on it for medication dosing
• use it for clinical practice or patient care
• attempt to reverse engineer or extract system behavior

7. Documents and Uploads

You may upload documents (e.g., lab reports, PDFs, records).

You represent you have the right to upload them.

Trellis processes these solely to provide services to you.

8. Data Protection

Trellis Intelligence operates inside a HIPAA-secured environment.

Your identifiable health information:

• is encrypted
• is not sold
• is not used for advertising
• is not used to train external AI models

See the Trellis Intelligence Privacy Policy for details.

Trellis also takes additional precautions to prohibit any 3rd party use, sharing or access to any de-identified health information for any commercial purpose. Your health data is not the business model - even de-identified.

9. No Medical Advice Disclaimer

The AI Services provide informational support only.

To the fullest extent permitted by law:

Trellis Health is not practicing medicine through the AI interface.

The AI output is not medical advice and should not be relied upon as a diagnosis or treatment recommendation.

This clause does not limit rights you have under healthcare or consumer protection law.

10. Limitation of Liability

You agree that Trellis Health is not liable for harm arising from reliance solely on AI responses where you did not seek appropriate clinical care.

Nothing in these terms limits liability that cannot legally be limited under applicable law.

11. Suspension

We may limit or suspend AI use if:

• misuse occurs
• safety risk is detected
• the system is being used for clinical decision-making outside its intended purpose

12. Changes

We may update the AI as it improves. Behavior and responses may change over time. Continued use constitutes acceptance of updated terms.

(“Trellis Intelligence Privacy Policy”)

Effective Date: Feb 12th 2026

This policy explains how Trellis handles data used by the Trellis Intelligence features, including conversational guidance, health summaries, document interpretation, and pattern recognition tools (collectively, the “Trellis Intelligence” or “AI Services”).

These terms are intentionally written in plain language to communicate transparently, if you have any questions please contact support@jointrellishealth.com.

By using Trellis Intelligence, you agree to the Trellis Intelligence Privacy Policy in addition to the Trellis Health General Terms of Service, Privacy Policy and Trellis Intelligence Terms.

If you do not agree, do not use the AI features.

1. Core Principle

Trellis Intelligence is a healthcare service, not an advertising platform.

Your health information is handled as Protected Health Information (PHI) under HIPAA. We do not monetize your data - even de-identified.

2. What the AI Receives

The system may process any or all of the following:

You provide:

• messages
• symptoms
• health questions
• personal context

You upload:

• lab reports
• PDFs
• medical records

Connected sources:

• clinical and medical records
• medications
• family history (if entered)
• connected wearable and environmental data

System generated:

• summaries and insights
• health patterns
• reminders

3. What We Never Do

Trellis does not:

• sell identifiable or de-identified health data
• use your conversations for advertising
• share data with data brokers, pharma or other data buyers
• allow external AI companies to train on your health information
• Trellis Health is not supported by insurance reimbursement and operates on a consumer cash-pay model. As such, no data is ever shared with your insurer, employer or government systems, without your active initiation.

Your conversations are not used to improve outside AI models.

4. AI Architecture & External Models

Trellis Intelligence operates on private, dedicated infrastructure within a HIPAA-secured Azure and AWS environments.

When machine-learning models are used:

• all queries are not linked with any identifiers or identifying information unless the user shares these in the chat
• conversation history is not transmitted with the query
• external models cannot retain your data
• no conversation history is shared
• queries are not sent to external models not hosted on dedicated infrastructure

Your medical records, conversation history, and memory remain encrypted inside the Trellis environment.

5. Training & Improvement

We may use internal, privacy-protected system data to:

• improve safety
• detect errors
• improve accuracy

This is performed within Trellis secure systems and not shared externally.

6. Human Review

Authorized staff may review limited content only when necessary for:

• safety investigation
• abuse prevention
• technical debugging

All access is logged and audited.

No employee may access records without a legitimate operational reason.

7. Clinician Access

A clinician sees your AI conversation only if you explicitly choose to share it or accept a connection initiated by you. The AI does not automatically send your messages to clinicians without your permission, except where required by law or imminent harm standards.

8. Safety Monitoring

The system may identify high-risk situations and initiate outreach offers.

However, Trellis is not a continuous monitoring or emergency response service.

9. Retention & Deletion

We retain data to provide your health timeline and continuity of care. Trellis Intelligence is GDPR compliant and you may delete your account at any time.

Upon deletion:

• your account is closed
• identifiable data is removed from active systems
• your personal health model is deleted

Encrypted backup copies may temporarily persist for legal and security compliance, after which they are purged.

10. Security

We use administrative, technical, and physical safeguards including:

• encryption at rest and in transit
• access control
• audit logging
• HIPAA security compliance

No system can be perfectly secure, but protecting your health data is a core operational obligation.

11. Children and Family Data

Guardians must have legal authority to manage dependent data.

Minors cannot independently consent to AI use.

12. Business Model Transparency

Trellis operates on subscription and care services revenue.

We do not rely on:

• advertising
• data sales, even de-identified
• behavioral profiling

Your information is used only to provide your healthcare services and insights.