Clinical Terms &
Conditions

Clinician Terms of Use,

Attestation, and Conditions for Safe Use

​

Last Updated: January 2026

By subscribing to and using Aion's clinical assessment support services (the "Service"), the subscribing regulated health professional (the "Clinician" or "Subscriber") acknowledges, agrees, and certifies the following terms with Aion HealthTech Inc. ("Aion").

This agreement operationalizes best practices for safe, transparent use of AI-supported tools in mental health care, including known risks such as over-reliance, automation bias, and potential effects on clinical skill maintenance.

 

​​

​

0. Definitions

 

AI-Assisted Output
Any summary, organization, structuring, suggestion, interpretation, or other output generated or supported by the Service using artificial intelligence or automated logic.

Clinician
A regulated health professional authorized under applicable law and professional standards to perform mental health assessment and clinical documentation within their scope of practice. At this time, the Service is deployed for use by physicians licensed to practice psychiatry and psychiatric nurse practitioners in Quebec, Canada, with expansion to other qualified mental health professionals subject to additional validation and regulatory authorization. The Clinician must use the Service only within their authorized scope of practice as defined by their professional licensing body.

High-Risk Scenario
A clinical context involving acute risk to patient or others, including but not limited to suicidality, self-harm, violence risk, severe instability, intoxication/withdrawal, psychosis, or altered mental status.

Original Patient Responses
Patient-provided inputs collected through the Service in their original or faithfully reproduced form.

Medical Record
The official clinical record maintained by or under the responsibility of the Clinician in accordance with applicable professional and legal requirements.

Safety Incident
Any event or concern suggesting the Service may have contributed to, or may contribute to, patient harm, clinical error, unsafe workflow, material misinformation, security/privacy compromise, or other clinically significant risk.
 

 

1. Regulatory Authorization, Clinical Evidence, and Transparency

​

1.1 Regulatory Status and Jurisdictional Context

At the time of execution of this Agreement, the Service holds the following regulatory status:

​

• Regulatory authorization/classification: Regulatory submission initiated and under preparation; pursuing Class II Software as a Medical Device authorization.

• Regulatory authority: Health Canada (Canada).

• Intended classification: Class II SaMD - to inform clinical management for serious mental health conditions.

• Manufacturer/Company ID: 193601

• Date of submission initiation: November 2025

​

The Service is not licensed or authorized for clinical use as a medical device at the time of execution of this Agreement and is being deployed under a pre-market regulatory pathway consistent with its current classification.

Provincial and Professional Regulatory Context

The Clinician acknowledges that use of the Service must comply with:

 

• Collège des médecins du Québec (CMQ) professional standards and deontological obligations,

• Quebec's Act respecting the protection of personal information in the private sector (Law 25),

• Quebec's health services and health information privacy requirements,

• Applicable CMQ guidance on AI-assisted clinical tools.

Jurisdictional Scope

The Service is intended for use by clinicians practicing in Quebec, Canada. Use in other jurisdictions requires confirmation of regulatory compliance in that jurisdiction.

Aion will make material changes in regulatory status available to subscribing clinicians through appropriate documentation or platform communication.

 

1.2 Intended Use Statement

The Service is intended to inform clinical management of adult patients (18+) presenting for psychiatric assessment by:

 

1. Collecting structured patient-reported information through digital assessment completed prior to clinical encounter;

2. Organizing and synthesizing patient responses into structured formats;

3. Presenting synthesized information to licensed mental health professionals authorized to perform psychiatric assessment for review and integration into diagnostic assessment and treatment planning.

The Service is a pre-assessment support tool. Clinical decisions are made by the Clinician following direct patient evaluation.
 

The Service is NOT intended to:

​

• Provide autonomous diagnosis or treatment recommendations,

• Replace psychiatric interview, mental status examination, or clinical judgment,

• Serve as an emergency intervention, crisis response, or triage system,

• Be used as the sole basis for clinical decisions in High-Risk Scenarios.

 

1.3 Clinical Validation and Evidence

The Clinician acknowledges that:

 

• At its current stage of development, the Service has not yet completed formal clinical validation studies intended to demonstrate clinical performance or effectiveness.

• Clinical evaluation activities to date have focused on design, usability, workflow alignment, risk analysis, and intended-use definition, consistent with early-stage development and pre-market regulatory preparation.

• The scope, design, and strength of clinical evidence will evolve over time based on regulatory classification, intended use, jurisdiction, and completion of planned validation activities.

Upon completion of clinical validation studies, Aion will make available to clinicians summary performance information including validation population characteristics, performance metrics as applicable to the validation design, known limitations by population or use case, and confidence intervals or uncertainty characterizations where available.

Aion maintains documentation, as applicable to the current development stage, including: intended use and risk analysis, system design and human-factors considerations, data governance and algorithm oversight documentation, and known limitations and assumptions.

 

1.4 Service Evolution and Change Management

The Clinician acknowledges that the Service may evolve over time through software updates, workflow refinements, and algorithmic changes.

Aion maintains internal change-management and algorithm-governance processes to assess the clinical relevance of such changes.

Where regulatory requirements apply, material changes affecting safety, performance, or intended use will be managed in accordance with applicable regulatory guidance.

 

1.5 Service Availability

Service availability commitments and technical failure procedures are defined in the Service Level Agreement, available separately.

 

1.6 No Misrepresentation

The Clinician agrees not to represent the Service to patients, colleagues, or third parties as having capabilities, authorization, or evidence beyond what is documented by Aion per Sections 1.1-1.3.

 

1.7 Regulatory Status Monitoring and Material Change Notification

Aion acknowledges that Clinicians have ethical and legal obligations to remain informed of regulatory actions, safety communications, and material changes affecting medical devices used in patient care.

​

​

1.7.1 Regulatory Actions Requiring Clinician Notification

​

Aion shall notify subscribing Clinicians, through written communication, in a reasonable timeframe upon becoming aware of any of the following events:

 

• Issuance of a regulatory warning letter, enforcement action, or formal notice of non-compliance by Health Canada, the FDA, or another competent regulatory authority;

• Suspension, revocation, or material condition imposed on a Health Canada medical device licence (once licensed);

• Mandatory recall, field correction, or safety notice issued or required by a regulator;

• Initiation of a regulatory investigation specifically concerning the Service's safety, performance, or compliance;

• Regulatory approval or requirement of a material modification requiring new authorization; or

• Voluntary withdrawal or suspension for safety or compliance reasons.

 

1.7.2 Safety Signals and Serious Adverse Events

Aion shall notify subscribing Clinicians in accordance with applicable regulatory requirements and internal safety protocols when Aion identifies:

 

• A patient death or serious injury potentially associated with Service use;

• A pattern of adverse events or safety signals suggesting a systematic risk; or

• Initiation of a corrective or preventive action related to such safety concerns.

 

 

1.7.3 Content of Notifications

Notifications shall include, as appropriate:

 

• A clear description of the issue or regulatory action;

• The clinical implications for patient care;

• Any recommended Clinician actions;

• The status and anticipated timeline for investigation or resolution; and

• Contact information for further questions or incident reporting.

 

1.7.4 Clinician Responsibilities Upon Notification

Upon receipt of a notification, the Clinician agrees to:

​

• Review the information promptly and assess whether continued use is appropriate for their patient population;

• Modify, suspend, or cease use as clinically and ethically indicated;

• Document receipt of the notification and any resulting action taken where relevant to patient care; and

• Notify affected patients where required by applicable law or professional obligations.

 

 

2. Conditions Precedent to Access

Access to the Service requires confirmation that:

 

2.1 Professional Liability Insurance and Practice Independence

Each Clinician maintains professional liability insurance appropriate to their scope of practice and jurisdiction, and confirms that their insurer is aware they use AI-assisted clinical tools.

Each Clinician practices under their own legal and commercial name, independently of Aion branding, and maintains their own professional identity separate from the Service provider.

 

2.2 Data Governance

Aion certifies that training datasets used for AI components do not contain identifiable personal health information and have been de-identified in accordance with applicable Canadian privacy legislation. Summary documentation of de-identification practices is available upon reasonable request.

 

2.3 Access to Patient-Provided and AI-Assisted Information

The Service provides the Clinician with access to patient-provided information and identifies which content is AI-Assisted Output versus patient-authored responses, enabling the Clinician to distinguish source and nature of information presented.
 

 

3. Scope of Use and Key Limitations

The Service shall be used only within the intended use described in Section 1.2. This Section defines the permitted scope of use, prohibited uses, and key limitations required for safe and responsible clinical use.

 

3.1 Permitted Scope of Use

Within its intended use, the Service may be used to support clinician-led mental health care workflows, including:
Collection of structured and narrative patient-reported information prior to or between clinical encounters;

 

• Organization and presentation of patient-provided inputs in a clinician-reviewable format;

• Generation of draft summaries or structured documentation support, provided that all content is independently reviewed, verified, and edited by the Clinician before incorporation into the Medical Record.

 

3.2 Prohibited and Non-Intended Uses

The Service is not intended to, and must not be used to:

 

• Replace clinical judgment, clinical examination, or professional responsibility;

• Function as diagnostic system, triage gatekeeper, crisis service, or treatment decision-maker;

• Prescribe, recommend, or determine medications, interventions, or care pathways without clinician oversight;

• Be relied upon as the sole or primary basis for high-risk clinical decisions, including decisions relating to emergency intervention, involuntary measures, or patient disposition; or

• Be represented to patients, colleagues, or third parties as having regulatory authorization, clinical indications, or validated performance beyond what is explicitly documented by Aion at the time of use.

 

3.3 General Limitations

The Clinician acknowledges that:

 

• Service outputs are dependent on patient self-report and contextual information and may be incomplete, inaccurate, or misleading;

• The Service may perform differently across clinical contexts, presentations, or patient populations; and

• The absence of an alert, flag, or highlighted content does not establish the absence of clinical risk.

• Accordingly, the Clinician shall apply independent clinical judgment and appropriate verification for all clinically meaningful interpretations and decisions.

 

3.4 High-Risk Scenario Restrictions

The Service is not intended for use as a primary assessment or decision-making tool in high-risk scenarios. Specific restrictions, contraindicated contexts, and required safeguards for high-risk scenarios are set out in Section 14 and form an integral part of the conditions of safe use under this Agreement.

 

 

4. Professional Responsibility and Clinical Authority

The Clinician acknowledges that the Service is a clinical support and data-structuring tool intended to support care delivery, documentation, and clinical reasoning, but not to diagnose, prescribe, or replace professional judgment.

The Clinician remains responsible for: clinical decision-making, diagnosis, treatment, referrals, follow-up, and ensuring care complies with applicable professional, ethical, and regulatory standards.

Aion remains responsible for: the design, development, maintenance, security, and regulatory governance of the Service within its stated intended use and documented limitations.

 

​

5. Patient Informed Consent and Transparency

The Clinician acknowledges that, in mental health care, informed consent for AI-assisted tools requires clear, plain-language disclosure of purpose, limitations, risks, and alternatives.

 

5.1 Required Patient Disclosure

Before providing a patient with access to the Service, the Clinician (or authorized clinical staff under the Clinician's direction) shall ensure the patient receives a plain-language explanation that includes, at minimum:

​​

1. Purpose and how the Service supports clinical assessment,

2. What information is collected and who processes it,

3. Risks and limitations (AI errors, privacy risks, clinician verification),

4. High-risk content workflow and crisis support availability,

5. Benefits of using the Service,

6. Alternatives and right to opt-out,

7. Right to withdraw consent at any time.

Patient Notification Workflow for High-Risk Content: The Clinician shall explain that the Service is a pre-assessment support tool and is not an emergency service. When the Service identifies potential high-risk content in patient responses, assessment results (including any risk indicators) are made available for clinician review. The Clinician must actively access and review the assessment before the clinical encounter. Clinical decisions regarding risk, safety, and disposition must be based on direct clinical evaluation.

A recommended plain-language disclosure template and documentation checklist is provided in Appendix A.

 

5.2 Capacity and Appropriateness to Consent

The Clinician shall use reasonable clinical judgment to assess whether the patient has capacity to provide informed consent to Service use.

If capacity concerns are apparent before assessment: defer Service use until capacity is established, or proceed with alternative assessment methods.

If capacity concerns arise during clinical evaluation: do not rely on assessment responses as a primary information source; conduct independent clinical assessment appropriate to the patient's current presentation; document the capacity concern and assessment approach used.

 

5.3 Documentation of Patient Consent

The Clinician shall document patient consent (or refusal/withdrawal) in the Medical Record, including at minimum:

​​

• Confirmation that the Service was explained in plain language,

• Whether the patient consented, declined, or withdrew consent,

• The date/time of consent (or withdrawal), and

• Any material concerns raised by the patient and how they were addressed.

Use of the template provided in Appendix A is strongly recommended to ensure consistent and auditable consent documentation.

 

5.4 Withdrawal and Continuity of Care

If a patient withdraws consent:

The Clinician shall cease using the Service for that patient going forward (subject to technical feasibility and record integrity),
Care shall continue using alternative documentation and assessment methods, and
Prior documentation created during the period of valid consent remains part of the Medical Record.

 

5.5 Multi-Clinician Workflows and Consent Scope

Patient consent for Service use is specific to the Clinician-patient relationship. Sharing Service-generated content with other clinicians requires separate patient authorization in accordance with applicable privacy law and professional standards.

 

5.6 Patient Access to Data and Deletion Rights

Patients have the right to access their assessment responses and AI-Assisted Outputs generated by the Service, and may request deletion of such data through mechanisms provided by Aion, in accordance with Law 25 and applicable privacy legislation.

The Clinician's Medical Record obligations are independent of patient requests to Aion for data deletion. Where clinically significant Service-generated content has been incorporated into the Medical Record, the Clinician follows applicable record retention requirements regardless of patient deletion requests to Aion.

 

 

6. Accuracy, Completeness, and Integrity of the Medical Record

The Clinician acknowledges that any Service-supported content incorporated into a patient record becomes part of the official Medical Record.

The Clinician is responsible for:

​​

• Reviewing, validating, correcting, and completing all Service-supported documentation prior to finalization; and

• Ensuring the final record accurately reflects the Clinician's own clinical assessment, reasoning, and conclusions.

When incorporating AI-Assisted Outputs into the Medical Record, the Clinician shall:

​​

• Independently verify accuracy and clinical appropriateness,

• Clearly document that AI-assisted tools were used, and

• Add the Clinician's own assessment and clinical reasoning.

AI-generated text may be edited and integrated as appropriate following verification.

The Clinician certifies that the Medical Record complies with applicable requirements for content, retention, confidentiality, access, and auditability, as defined by law, professional orders, and institutional policies.

 

 

7. Use of AI: Supportive, Non-Diagnostic Role

​
7.1 Review of Patient-Provided and AI-Assisted Content

The Clinician shall review both patient-provided information and AI-Assisted Outputs as described in Section 2.3, with particular attention to original patient-authored content in:

​​

• High-Risk Scenarios (Section 14)

• Cases where AI synthesis may be incomplete or uncertain

• Complex or atypical presentations

 

7.2 AI-Assisted Outputs as Supportive Tools

The Clinician acknowledges and agrees that:

​​

• AI-Assisted Outputs are provided to support clinical work and must be critically appraised; and

• The Service is not used for autonomous diagnosis or autonomous treatment determination.

The Clinician shall not represent the Service to patients or third parties as a replacement for professional judgment or as an autonomous diagnostic authority.

 

7.3 AI Output Transparency and Access to Documentation

The Parties acknowledge that safe use of AI-assisted tools requires access to information that supports appropriate interpretation, including limitations, provenance, and (where available) uncertainty indicators.

 

7.3.1 Output-Level Transparency

For each AI-Assisted Output, Aion will make available information sufficient for the Clinician to understand:

​

1. Output type and purpose;

2. Source labeling identifying AI-Assisted Output versus patient-authored content;

3. Version identification sufficient to link the output to the Service version in use; and

4. Known limitations relevant to that output type.

Aion will display key limitations and warnings directly within the Service interface at the point of use.

 

7.3.2 Uncertainty and Confidence Indicators

Where the Service provides confidence/uncertainty indicators, they will be displayed in a manner designed to support clinical judgment and avoid false precision.

The Clinician acknowledges that absence of an uncertainty display does not imply high certainty, and that AI-Assisted Outputs must be interpreted in clinical context.

 

7.3.3 Transparency Documentation

Aion maintains and makes available to Clinicians:

Readily accessible:

​

• Current limitations and warnings

• Regulatory status and safety communications (Section 1.7)

• Material update notifications

Upon reasonable request (subject to confidentiality constraints):

​

• System description and intended use documentation

• Validation status summary (Section 1.3)

• Data governance summary (Section 2.2)

When validation metrics become available, Aion will display them prominently in Clinician-facing materials.

 

7.3.4 High-Risk Outputs and Crisis-Related Content

For any feature that may contribute to High-Risk Scenario identification, the Service will provide clear messaging that:

​​

• The tool is supportive and does not replace direct assessment;

• Absence of an indicator does not establish absence of risk; and

• The Clinician must apply independent clinical judgment and follow crisis protocols as required.

 

​

8. Human-AI Interaction and Automation Bias

The Clinician acknowledges that literature in medicine identifies potential unintended effects of AI-supported tools on clinical practice, including over-reliance, reduced independent verification, and potential effects on clinical skill maintenance.

Accordingly, the Clinician agrees that:

​

• The Service must not be used as a substitute for maintaining or exercising professional competence;

• AI-Assisted Outputs must be actively reviewed, questioned, and contextualized; and

• The Clinician remains responsible for preserving clinical skill, judgment, and safe practice patterns over time.

 

​

9. Training, Ongoing Competence, and Clinical Skill Maintenance

The Parties acknowledge that AI-assisted clinical tools may, over time, influence clinician reasoning patterns. Responsible use therefore requires initial training, ongoing competence maintenance, and active attention to clinical skill preservation.

 

9.1 Initial Training and Attestation

Prior to using the Service in clinical practice, the Clinician certifies completion of Aion's onboarding training, which covers, at minimum:

​​

• The Service's intended use, non-intended use, and limitations;

• High-Risk Scenarios and contraindicated contexts (Section 14);

• Documentation and auditability requirements (Section 12);

• AI-specific risks, including automation bias and over-reliance; and

• Appropriate integration into human-led clinical workflows.

The Clinician attests that they understand the vulnerabilities of the mental-health populations they serve and that the Service does not replace independent diagnostic reasoning or clinical judgment.

 

9.2 Ongoing Competence and Continuing Education

The Clinician agrees to maintain ongoing familiarity with the Service through:

​

• Review of material product changes, safety communications, and updated limitations per Sections 1.7 and 9.4; and

• Completion of required training modules as specified by Aion following material changes affecting safe use.

Aion may require completion of targeted refresher training where a material safety, bias, or performance issue is identified.

 

9.3 Maintaining Clinical Skills and Professional Judgment

The Clinician commits to:

​​

• Performing independent clinical assessment and diagnostic reasoning alongside Service use

• Using the Service as a pre-assessment tool that supports (not replaces) direct psychiatric evaluation

• Seeking consultation, supervision, or additional training if concerns arise about clinical decision-making

If the Clinician identifies concerns that use of the Service may be affecting clinical reasoning or confidence, the Clinician may increase independent review, seek peer consultation, request additional training from Aion, or temporarily adjust Service use while addressing the concern.

 

​

9.4 Material Algorithm Updates and Change Governance

Aion acknowledges that uncontrolled or opaque changes to AI-assisted systems may pose clinical safety risks. Accordingly, Aion commits to maintaining an algorithm change governance framework appropriate to the Service's regulatory classification and stage of development.

 

9.4.1 Definition of Material Algorithm Update

A material algorithm update is defined as any change that may reasonably affect:

​​

• Performance characteristics relevant to clinical use;

• Intended use, scope, or stated limitations;

• Handling of High-Risk Scenarios;

• Safety features, risk thresholds, or alerting logic;

• Contraindications or recommended use limitations.

Non-material updates (e.g., bug fixes, UI changes, performance optimizations not affecting clinical logic) may be deployed without advance notice.

 

9.4.2 Notification of Material Updates

Aion will notify subscribing Clinicians of material algorithm updates in a reasonable timeframe in accordance with applicable regulatory requirements and internal governance policies.

Such notification will include:

​​

• A plain-language summary of the change and its rationale

• Expected impact on system behavior

• Any new or revised limitations, warnings, or use considerations

• Reference to any required training per Section 9.2

 

9.4.3 Clinician Review and Adoption

The Clinician may defer adoption of a material algorithm update for a reasonable period, where technically feasible, to allow for review of provided information.

Where the Clinician reasonably believes that a material update raises safety or appropriateness concerns for their practice context, the Clinician may temporarily suspend use without penalty and notify Aion via the incident or feedback mechanisms described in Section 15.

 

9.4.4 Emergency Updates

In the event of an urgent safety issue requiring immediate action, Aion may deploy emergency updates without advance notice. In such cases, Aion will notify Clinicians as soon as practicable and provide guidance on any interim safety measures required.

 

9.4.5 Version Identification

Aion will make the current Service version identifier available within the Service. Documentation considerations for version tracking are addressed in Section 12.3.

 

 

10. Accounts, Authorized Users, and No Credential Sharing

Each user must have a unique account. The Clinician shall not share credentials or permit unauthorized access.

If the Clinician permits supervised staff or delegates to use the Service under the Clinician's authority (where permitted), the Clinician remains responsible for ensuring such users are appropriately authorized, trained, and operating under appropriate supervision and role-appropriate permissions.

 

 

11. Transparency Toward Patients

The Clinician agrees to use the Service transparently, including:

​​

• Informing patients, where appropriate and required, that AI-assisted tools support documentation or assessment; and

• Respecting applicable consent, disclosure, and information obligations.

 

 

12. Documentation Requirements

The Clinician acknowledges that appropriate documentation of Service use is necessary to support continuity of care, patient safety, and medico-legal auditability.

 

12.1 Basic Documentation (All Service-Assisted Encounters)

When the Service is used to support clinical evaluation, the Medical Record shall include:

​​

1. Notation that the Service was used

2. The Clinician's clinical assessment, diagnosis, and treatment plan based on independent clinical evaluation

 

12.2 AI-Assisted Content in Medical Records

If AI-Assisted Outputs are incorporated into the Medical Record, they must be:

​​

• Clearly labeled as AI-assisted

• Independently verified by the Clinician

• Accompanied by the Clinician's clinical assessment and reasoning

 

12.3 High-Complexity Cases (Discretionary)

For contested diagnoses, medico-legal concerns, or significant discrepancies between Service synthesis and clinical assessment, the Clinician may document additional detail including:

​​

• Service version identifier (available per Section 9.4.5)

• Specific outputs reviewed and clinical rationale for deviations

• Additional verification steps taken

 

12.4 Service Audit Trail

Aion maintains system logs appropriate to support safety review and incident investigation, including records of Service use, system version, and patient assessment data, in accordance with Aion's Data Retention and Deletion Policy.

These are available to the Clinician upon reasonable request for patient safety review, incident investigation, or medico-legal purposes, subject to applicable privacy constraints and role-based permissions.

 

12.5 Record Retention

The Clinician remains responsible for Medical Record retention per applicable law and professional obligations.

Aion's data retention practices are described in Aion's Data Retention and Deletion Policy, available upon request. The Clinician acknowledges they have reviewed or had the opportunity to review this policy.

 

​

13. Sharing Information With Other Providers

When sharing Service-related information with other care providers involved in the patient's care, the Clinician shall preserve transparency by:

​

• Sharing both patient-provided information and AI-Assisted Outputs where clinically appropriate, and

• Clearly distinguishing AI-assisted content from the Clinician's clinical conclusions.

 

 

14. High-Risk Scenario Management and Restrictions on Use

 

14.1 Core Principles for High-Risk Contexts

The Clinician acknowledges that mental health contexts involving suicide risk, self-harm, violence risk, severe instability, intoxication/withdrawal, psychosis, delirium, or other acute safety concerns require immediate human clinical oversight and may require emergency intervention.

The Service is a pre-assessment support tool, not an emergency response, crisis intervention, or monitoring system. Time-sensitive safety decisions must be based on direct clinical evaluation.

​​

1. Pre-Assessment Function: The Service collects patient-reported information before clinical evaluation. High-risk content may be present in patient responses. The Clinician is responsible for reviewing assessment content and conducting appropriate clinical evaluation.

2. Clinician Review Responsibility: Clinicians are responsible for actively reviewing assessment results before clinical encounters. The Service may provide notifications when high-risk content is identified, but such notifications do not replace the Clinician's obligation to review and assess directly.

3. Direct Clinical Assessment Required: Clinical decisions regarding risk, safety, and disposition must be based on direct clinical evaluation, not solely on assessment responses.

4. Automated Indicators as Supportive Only: The Service may include risk indicators. The presence or absence of an indicator does not determine clinical risk and does not replace clinical assessment.

 

14.2 Suicidal Ideation, Self-Harm, and Acute Risk

When the Clinician reviews assessment content and identifies potential self-harm, suicide risk, risk of harm to others, or severe instability, the Clinician shall:

​​

• Perform timely direct clinical assessment (in-person or synchronous telehealth as clinically appropriate)

• Apply standard-of-care risk assessment and crisis management procedures

• Determine appropriate disposition based on independent clinical judgment

• Document that high-risk content was identified and that independent clinical assessment and disposition were completed

 

14.3 Crisis Contexts Requiring Enhanced Clinical Primacy

Service-assisted information should not be the PRIMARY basis for clinical decision-making in active crisis situations including:

​​

• Active suicidal or homicidal crisis requiring immediate intervention

• Acute intoxication or withdrawal

• Severe psychosis, delirium, or markedly altered mental status

• Imminent psychiatric hospitalization

In such contexts, the Clinician shall:

​​

• Conduct direct clinical assessment as the primary basis for decisions

• Use assessment responses only as supplementary background information

• Apply heightened independent verification per Section 14.1

• Follow standard-of-care crisis protocols

 

14.4 Pediatric Use / Age Restrictions

Unless otherwise specified in Aion's intended use documentation, the Service is intended for patients 18 years and older.

If the Service is inadvertently used with a patient under 18, the Clinician shall:

​​

• Discontinue Service use for that patient

• Complete independent clinical assessment to ensure no harm resulted

• Report to Aion through standard incident reporting (Section 15)

 

14.5 Patients With Disabilities (Cognitive, Developmental, Communication, Sensory)

Before relying on patient self-report collected via the Service, the Clinician agrees to consider whether the patient's disability or communication needs may affect ability to understand and meaningfully answer questions, reliability of self-report without accommodation, or ability to provide informed consent.

Where significant impairment is present, the Clinician agrees to use the Service only with appropriate safeguards, which may include:

​​

• Involvement of a caregiver/guardian or informant input where clinically appropriate and lawful;

• Additional clinical interviewing and corroboration;

• Use of accessible formats or accommodations as available and appropriate; and

• Enhanced independent verification of any AI-Assisted Output.

 

14.6 Rare, Atypical, or Low-Prevalence Presentations

For rare, atypical, or complex presentations outside the Service's documented scope, limitations, or validated use cases (as available), the Clinician agrees not to rely on the Service as a primary basis for clinical decisions and to apply enhanced independent verification, including specialist consultation where appropriate.

 

​

15. Safety Incident Reporting and Clinical Disagreement

The Clinician agrees to report Safety Incidents and clinically significant disagreements with AI-Assisted Outputs through Aion's designated reporting mechanism.

 

15.1 Critical Safety Incidents

Report as soon as reasonably practicable:

​​

• Patient death potentially associated with Service use

• Serious patient harm potentially associated with Service use

• Major security breach affecting patient data

• System failure preventing access in time-sensitive situation

 

15.2 Serious Safety Incidents

Report in a reasonable timeframe:

​​

• Material inaccuracy in AI-Assisted Output affecting clinical decisions

• Systematic performance issues or concerning patterns

• Suspected algorithmic bias affecting patient care

• Service malfunction in High-Risk Scenario, including:

  • False negative: Service failed to surface high-risk content clearly present in patient responses

  • Systematic false positives causing patient harm or distress

  • Technical failures affecting access to risk assessment information

 

15.3 Clinical Disagreement Reporting

When the Clinician's assessment significantly differs from AI-Assisted Output in ways that could indicate algorithmic error, bias, or systematic performance issues, the Clinician should report this through Aion's reporting mechanism with:

​

• Brief description of the discrepancy

• Clinical rationale for the disagreement

• Approximate patient presentation context (de-identified)

Such reporting supports ongoing safety monitoring and algorithm improvement and does not imply Service malfunction.

 

15.4 Non-Urgent Concerns

Report in a reasonable timeframe:

​​

• Usability issues

• Minor output inconsistencies

• Feature requests or general feedback

Important clarification: Routine identification of suicidal ideation or risk factors through proper Service functioning is the intended use and does NOT require reporting. Only Service malfunctions or safety concerns require reporting under this Section.

 

15.5 Confidentiality in Reporting

The Clinician shall report in a manner that:

​

• Respects patient confidentiality and privacy laws

• Minimizes personal information shared (de-identify when appropriate)

• Complies with professional secrecy obligations

Where required by law or professional obligations, affected patients will be informed.

 

 

16. Cybersecurity and Secure Use Responsibilities

The Clinician agrees to use the Service in a secure manner, including:

​​

• Using reasonably secured devices and networks,

• Applying reasonable security hygiene (e.g., updates, device protection),

• Enabling available security features (such as multi-factor authentication if offered), and

• Promptly reporting suspected credential compromise, unauthorized access, or device loss where it could affect patient data.

 

​

17. Bias, Fairness, and Post-Market Monitoring

The Parties acknowledge that responsible deployment of AI-assisted tools in mental health requires proactive monitoring for safety signals, performance degradation, unintended consequences, and inequitable outcomes across patient groups.

 

17.1 Aion's Commitments

Aion maintains a post-market surveillance program to monitor for safety signals, performance issues, bias, and inequitable outcomes. Implementation details are provided in Appendix B.

Aion will notify Clinicians when material safety signals, limitations, bias concerns, or corrective actions are identified that affect clinical use. Such notifications will include, as appropriate:

​​

• A plain-language summary of the issue

• Potential clinical implications

• Recommended clinician actions

• Status and next steps

 

17.2 Clinician Responsibilities

The Clinician agrees to:

​​

• Report suspected Safety Incidents per Section 15

• Report suspected bias or inequitable performance concerns where observed

• Cooperate reasonably with safety investigations by providing de-identified information where lawful and feasible

• Complete updated training where material safety or fairness issues require changes in safe use practices

 

 

18. Liability Allocation, Indemnification, and Insurance

 

18.1 Shared Responsibility Framework

Safe use of the Service depends on both Parties fulfilling their respective obligations under this Agreement. Aion is responsible for Service design, maintenance, security, and compliance within documented limitations. The Clinician is responsible for clinical decision-making, patient consent, and appropriate use within intended scope and limitations.

The following indemnification provisions allocate liability consistent with these responsibilities.

 

18.2 Aion Indemnification (Third-Party Claims)

To the extent permitted by applicable law, Aion shall defend, indemnify, and hold harmless the Clinician and the Clinician's practice/clinic (if applicable) from and against third-party claims, damages, losses, and reasonable costs and expenses (including reasonable legal fees) arising out of or relating to:

​​

1. Aion's breach of privacy/security obligations, including security incidents attributable to Aion's failure to implement reasonable administrative, technical, and physical safeguards for the Service;

2. Aion's material breach of this Agreement relating to transparency commitments, algorithm update notifications (Section 9.4), or de-identification and data governance assurances (Section 2.2);

3. Aion's gross negligence or willful misconduct in the development, maintenance, or operation of the Service;

4. Intellectual property infringement claims alleging that the Service, when used as authorized, infringes a third party's intellectual property rights.

 

18.3 Indemnification Exclusions

Aion's indemnification obligations under Section 18.2 shall not apply to the extent a claim arises from:

​​

1. Use of the Service in contraindicated emergency/crisis contexts where the Service was the PRIMARY basis for clinical decisions (per Section 14.3)

2. Material failure to obtain patient consent where legally required

3. Intentional misrepresentation of Service capabilities to patients

4. Unauthorized modification of software

5. Willful violation of security requirements

For clarity: Good-faith clinical judgment calls do not void Aion's indemnification provided the Clinician exercised reasonable care and independent clinical assessment.

 

18.4 Clinician Indemnification (Limited)

The Clinician shall indemnify and hold harmless Aion from third-party claims arising out of the Clinician's use of the Service outside the permitted scope as defined in Section 18.3, or the Clinician's willful misconduct or gross negligence in clinical care, to the extent permitted by law.

 

18.5 Limitation of Liability

Except for liability that cannot be limited by law, and except for Sections 18.2 (Aion indemnities), gross negligence, willful misconduct, and breaches of confidentiality or security obligations, each Party's aggregate liability arising out of this Agreement shall be limited to an amount equal to the total fees paid (or payable) by the Clinician to Aion for the Service in the twelve (12) months preceding the event giving rise to liability.

 

18.6 Aion Insurance

Aion shall maintain, at its own expense, commercially reasonable insurance coverage appropriate to the nature of the Service, including the development and operation of a digital health platform processing sensitive mental health information.

Such coverage shall include, as applicable:

​​

1. Professional Liability/Technology Errors & Omissions;

2. Network Security and Privacy Liability;

3. Commercial General Liability.

All policies shall be maintained with reputable insurers and shall be consistent with industry standards for health technology and Software as a Medical Device (SaMD) providers.

Upon reasonable request, Aion shall provide the Clinician with certificates of insurance evidencing that such coverage is in force.

Aion shall notify subscribing Clinicians of any material reduction or non-renewal of the above coverage where reasonably practicable.

 

18.7 Clinician Insurance

The Clinician remains responsible for maintaining professional liability insurance appropriate to their scope and jurisdiction, and for confirming with their insurer that coverage extends to AI-assisted documentation and workflow tools used as part of clinical care.

 

 

19. Term, Termination, and Transition

 

19.1 Subscription Term

This Agreement remains in effect while the Clinician maintains an active subscription to the Service.

 

19.2 Termination by Clinician

The Clinician may terminate their subscription at any time by providing thirty (30) days written notice to Aion.

 

19.3 Termination by Aion

Aion may terminate this Agreement:

​​

• With sixty (60) days notice for convenience

• Immediately for material breach (with opportunity to cure where patient safety is not at risk)

• Immediately for non-payment after thirty (30) days grace period

 

19.4 Data Access Upon Termination

Upon termination:

​​

• The Clinician retains access to historical patient data for sixty (60) days to facilitate data export and transition

• Aion will provide data export in standard formats upon request

• Following the transition period, data retention follows Aion's Data Retention and Deletion Policy

• The Clinician remains responsible for Medical Record retention obligations independent of Aion's retention

 

19.5 Service Discontinuation

If Aion ceases operations or discontinues the Service:

​​

• Aion will provide ninety (90) days advance notice where reasonably practicable

• Aion will facilitate data export and transition assistance

• Clinician notification obligations (Section 1.7) remain in effect through the discontinuation period

 

19.6 Material Changes to This Agreement

Material changes to this Agreement (affecting clinical use obligations, liability allocation, or regulatory commitments) will be communicated to subscribing Clinicians at least thirty (30) days before taking effect, with opportunity to review and terminate if the changes are unacceptable.

Non-material changes (clarifications, formatting, administrative updates) may be implemented with notice but without advance review period.

Continued use of the Service following notice of material changes constitutes acceptance.

 

 

20. Governing Law, Jurisdiction, and Dispute Resolution

 

20.1 Governing Law

This Agreement shall be governed by and construed in accordance with the laws of the Province of Quebec and the federal laws of Canada applicable therein, without regard to conflict of law principles.

 

20.2 Jurisdiction

The Parties irrevocably submit to the exclusive jurisdiction of the courts of the Province of Quebec for any dispute arising out of or relating to this Agreement.

 

20.3 Language

The Parties confirm their wish that this Agreement and all related documents be drawn up in English. Les parties confirment leur volonté que la présente convention ainsi que tous les documents s'y rattachant soient rédigés en anglais.

 

 

Appendix A - Patient Consent Documentation Template

PATIENT CONSENT FOR AI-ASSISTED ASSESSMENT

Date: ____ / ____ / ____ Time: ______

Patient informed that AI-assisted assessment tool is used to organize patient-provided information for clinician review.

☐ Patient consented to AI-assisted assessment for support in care
☐ Patient declined (standard workflow used)
☐ Patient withdrew consent (AI-assisted assessment discontinued going forward)

Patient informed of:

☐ Purpose and how the tool works (support tool; clinician decides)
☐ Risks/limitations (possible errors; privacy risks)
☐ Mental health safety limits (not emergency/crisis tool; clinician will assess directly)
☐ Right to opt out / withdraw without impact on care

Capacity check:

☐ Patient appears capable of informed consent at time of discussion
☐ Capacity in question → Assessment deferred / alternative process used

Clinician signature/initials: _____________________

 

​

Appendix B - Post-Market Surveillance Program

Aion's post-market surveillance program includes the components described in Section 17, implemented through:

Centralized Tracking System: All safety incidents, complaints, and concerns reported under Section 15 are tracked centrally with periodic trend analysis.

Performance Monitoring Methods: Monitoring includes tracking output pattern changes, user override rates (where logged), incident patterns, and deviations from expected behavior. Where formal validation baselines are not established, risk-based operational indicators guide investigation.

Fairness Monitoring Approaches: Where demographic data collection is restricted, ethically appropriate proxy-safe analyses and targeted consented evaluation initiatives are used.

Investigation and CAPA Procedures: Issues are triaged by severity. Serious safety concerns are investigated as soon as practicable. Corrective actions may include remediation, updated warnings, workflow changes, algorithm adjustments, or feature restrictions.

Regulatory Cooperation: Aion meets applicable reporting obligations and cooperates with health authorities per Section 1.7.

​

​

END OF DOCUMENT