Conversational AI & mental health

AI. AI. AI.

Be AWARE. Beware.
Scientific source & licence

Based on the 2026 AWARE framework article

This page is an educational adaptation created by Augmenta Health to make the framework easier to understand and use in reflection, conversation and professional practice. It is not presented as an official derivative endorsed by the article authors.

Hudon A, Bhat V, Nunez J-J, Paquin V, Malouin-Lachance A, Lespérance F.
The AWARE (AI Use, Why, Attachment, Reality and Risk, and Effect on Functioning) Framework for Assessing Patients’ Use of AI in Mental Health Care.
JMIR Medical Education. 2026;12:e108766. DOI: 10.2196/108766 · Original article: JMIR Medical Education.

Original work © Alexandre Hudon, Venkat Bhat, John-Jose Nunez, Vincent Paquin, Amylie Malouin-Lachance, François Lespérance. Originally published in JMIR Medical Education, 16 September 2026. The article is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence, which permits reuse and adaptation with appropriate attribution.

Conversational AI is becoming part of how people think, decide, learn, cope, seek reassurance and sometimes look for emotional support. The AWARE framework offers five areas to help people reflect on their own AI use and help professionals understand the role AI may be playing in someone’s life.

Watch the overview ↓Explore AWARE

A flexible educational and interviewing framework. Not a diagnostic, screening or risk-prediction tool.

Start here · Under 1 minute

What is the AWARE framework?

One short overview introduces the five areas: AI Use, Why, Attachment, Reality & Risk, and Effect on Functioning.

Use AWARE as a way to notice, ask and understand the role conversational AI may be playing — in your own life or in a conversation with someone else.

The same five domains can support personal reflection and professional conversations. The framework does not produce a score or replace established clinical assessment.

Start here · AWARE overview

Watch on YouTube
Why this matters now

AI is not just another search box.

Search engines mainly help us find information. Conversational AI can respond, adapt, continue a dialogue and produce language that may feel validating, supportive or emotionally meaningful.

From tool → to conversational environment

People may use conversational AI for education, productivity, health information, self-reflection, decisions, emotional support or companionship. That makes the pattern and meaning of use relevant — not merely the fact that someone uses AI.

“Feeling understood can be psychologically meaningful — even when the understanding is generated by a machine.”

The point is not to dismiss the experience, nor to confuse generated responsiveness with reciprocal human understanding.

AAI Use

What, where, how often — and what is being shared?

 WWhy

What need or purpose is bringing someone to AI?

 AAttachment

What emotional or relational meaning has the interaction acquired?

 RReality & Risk

Is AI influencing beliefs, decisions, treatment or safety?

 EEffect on Functioning

What is changing in sleep, work, relationships, care or recovery?
The AWARE framework

Five areas to notice, explore and talk about.

Use them for personal reflection or as flexible conversation prompts in clinical and helping contexts. They are not a test and do not replace established assessment.

AI Use

What is the pattern of AI use — and what is being shared?

Explore: Start with the practical footprint: which tools are used, how often, in what situations, through which modes, and what personal or health information is shared. Using AI is not itself a problem; the pattern, purpose and impact give it meaning.

01 · AI Use

Watch on YouTube

Look at the pattern

  • Platform or type of AI
  • Frequency, duration and intensity
  • Text, voice, images or uploaded documents
  • Occasional use versus repeated or exclusive use

Look at privacy

  • Symptoms or mental health history
  • Medication or treatment information
  • Medical reports, images or identifiers
  • Information about family or other people
What do you use AI for, and how often?
What kinds of personal or health information do you share with AI?
Have you uploaded reports, images or other documents?
How do you decide what is appropriate to share with an AI system?
AI use is context, not a diagnosis. Look at the pattern, purpose, information shared and effect on everyday life.

Why

What need or purpose is bringing you to AI?

Explore: Ask what brings someone back. Learning, convenience, creativity, decision support, reassurance, loneliness and emotional support may look similar from the outside but can meet very different needs.

02 · Why

Watch on YouTube

Common purposes

  • Learning or productivity
  • Creativity and problem-solving
  • Health information
  • Decision support

When more context may help

  • Emotional regulation
  • Loneliness or companionship
  • Repeated reassurance seeking
  • Coping with distress
What do you find most helpful about AI?
When do you choose AI instead of another person or resource?
What role does AI play when you are stressed or upset?
What need does AI meet that other resources do not?
A useful question is not simply “How much AI?” but “What is AI doing for this person — and what happens around that use?”

Attachment

What emotional or relational meaning has AI acquired?

Explore: Conversational AI can feel available, attentive, reassuring or understanding. Explore whether the interaction has become emotionally meaningful, trusted, relied upon, or more appealing than support the person actually wants from others — without assuming that attachment itself is pathological.

03 · Attachment

Watch on YouTube

Different experiences

  • Trust
  • Companionship
  • Feeling understood
  • Emotional reliance

Do not collapse these together

Therapeutic alliance, anthropomorphism, parasocial interaction, attachment and dependence are related ideas, but they are not the same thing.

How would you describe your relationship with AI?
What is it like when you cannot use it?
Do you feel understood when interacting with AI?
How do you decide when to trust AI and when to trust people?
Feeling connected to an AI system is not automatically evidence of illness. Explore the nature, intensity and functional significance of the relationship.

Reality & Risk

Is AI influencing what feels true, important or safe?

Explore: Look more closely when AI begins shaping important beliefs, self-understanding, treatment or medication decisions, major life choices, or what someone does during a crisis. Positive findings require established clinical assessment beyond AWARE.

04 · Reality & Risk

Watch on YouTube

Reality & judgement

  • Self-identity or fixed beliefs
  • Major life decisions
  • What is considered accurate or trustworthy
  • Whether AI is being used to decide what is “real”

Risk & safety context

  • Suicidal thoughts or self-harm
  • Violence
  • Treatment refusal or medication changes
  • Unsafe health decisions
Has AI influenced how you understand yourself or an important situation?
Do you use AI to decide whether something is true or real?
Has AI influenced treatment, medication or another major decision?
Have you turned to AI during a crisis? What happened next?
Important: AWARE does not assess suicide, self-harm or violence risk. Any concern in these areas requires established clinical risk assessment independent of AWARE.

Effect on Functioning

What has changed in everyday life since AI became part of it?

Explore: Bring the question back to ordinary life. AI may help with learning, organisation, coping or support, while also affecting sleep, work or study, relationships, routines, treatment engagement or recovery. Look for both benefits and difficulties.

05 · Effect on Functioning

Watch on YouTube

Check everyday life

  • Sleep
  • Work or study
  • Relationships
  • Daily routine and quality of life

Check care & recovery

  • Coping
  • Treatment engagement
  • Help-seeking
  • Social connection or withdrawal
Has AI changed your daily routine or sleep?
Has it changed how you work or study?
Has it changed your relationships?
Has it changed how you engage with mental health care or recovery?
The aim is balance: understand whether AI is helping, interfering, or simply changing how someone functions day to day.
Keep the frame balanced

AI may help. AI may complicate. Sometimes both are true.

The aim is not to fear AI or idealise it. Look at the pattern, purpose, context and effect — including both benefits and concerns.

Possible helpful roles

  • Psychoeducation and easier access to information
  • Learning, problem-solving and self-reflection
  • Coping strategies or emotional expression
  • Structured mental health support in some purpose-built systems
  • Support between or alongside professional care in appropriate contexts

Reasons to look more closely

  • Repeated reassurance seeking or emotional reliance
  • Replacement of wanted human relationships or supports
  • AI reinforcing fixed or unsafe beliefs
  • AI influencing treatment or medication decisions
  • Excessive use that affects sleep, work, relationships or recovery
A simple AWARE pause

Five questions to keep in mind.

Use them for yourself or in a conversation with someone else. They are prompts for reflection and understanding — not a test, diagnosis or score.

AWhat is AI being used for — and what is being shared?
WWhy is AI being used at this moment?
AWhat emotional meaning has the interaction acquired?
RIs AI influencing important beliefs, decisions or safety?
EWhat effect is AI having on everyday life?
Notice. Ask. Understand the context.
For clinical and helping conversations

The same five domains can organise a brief AI history.

For professionals, AWARE can act as a flexible conversation map when AI use is relevant. It complements — and does not replace — existing psychiatric interviewing, clinical formulation or risk assessment.

The 60-second AI history

  1. Use: “Do you use conversational AI such as ChatGPT, Gemini, Claude or another chatbot? What for?”
  2. Why: “When are you most likely to turn to it?”
  3. Attachment: “Does it ever feel emotionally important, reassuring or like a source of understanding?”
  4. Reality & Risk: “Has it influenced important beliefs, treatment decisions, medication, help-seeking or crisis decisions?”
  5. Effect: “Has it changed sleep, work, study, relationships or recovery?”

Document the role — not merely the app.

“Uses ChatGPT” alone tells you little. When clinically relevant, document pattern, purpose, psychological significance, impact and any benefits or concerns.

Example structure
Uses conversational AI regularly for emotional support and problem-solving; describes feeling understood; no reported AI-related change in major beliefs or treatment decisions; reports improved coping and no perceived impairment in work, social or academic functioning. Separate standard risk assessment if clinically indicated.
Where could AWARE fit in education?

The source article proposes possible use in undergraduate and postgraduate teaching, simulation, case-based learning, OSCEs, clinical supervision, workplace-based assessment, reflective practice and continuing professional development. These are educational applications, not evidence that AWARE itself is clinically validated.

Is AWARE a validated clinical tool?

No. The authors describe it as an author-proposed conceptual interviewing framework. It has not undergone formal psychometric development or validation; its five domains should be treated as provisional and subject to content validation, feasibility testing, educational evaluation and cross-cultural adaptation.

If AI becomes part of a crisis, bring established human safety assessment back into the picture.

AWARE can help identify that AI has become part of the context. It does not replace assessment of suicide, self-harm, violence, impaired reality testing or other acute psychiatric risk.

Immediate support in India: Tele-MANAS is a Government of India 24×7 tele-mental-health service available at 14416 or 1800-89-14416. In an immediate emergency, use local emergency services. This page is educational and is not a crisis service.
Use the framework carefully

What AWARE is — and what it is not.

AWARE is

A flexible educational and interviewing framework for exploring how conversational AI may intersect with mental health, functioning, relationships, treatment and safety.

AWARE is not

A diagnostic test, screening tool, psychometric scale, risk-prediction instrument, universal checklist or substitute for standard psychiatric assessment.

The takeaway

AI can inform. AI can support. AI can influence.

Understand the role it is playing.

Be AWARE. Beware.