A diagnosis can change a life.

It can give language to experiences that were previously dismissed, misunderstood, or blamed on character.

It can help a family understand why a child communicates, moves, regulates, learns, or responds differently.

It can open access to services, accommodations, education plans, therapies, assistive technology, funding, and legal protection.

It can help a person recognise that years of exhaustion were not simply failure.

It can also become a trap.

A diagnosis can be treated as though it explains everything.

A support level can become a permanent identity.

A professional category can begin to outrank the person’s present expression.

A child can enter a system through one diagnostic doorway and spend the rest of life being spoken to through that doorway.

A person who has no diagnosis may be treated as though their needs are not real.

A person who masks successfully may be denied support because their difficulty is not visible enough.

A person with high support needs may be denied opportunity because dependence is mistaken for absence of ability.

A person with intellectual disability may be erased by optimistic claims that every difficulty is merely hidden competence.

A diagnosis is useful.

But a diagnosis is not a person.

That distinction must be built into Amanah Companions from the beginning.

Diagnosis has a legitimate place

Amanah should not be anti-diagnosis.

Autism diagnosis is a professional process based on developmental history and observed behaviour. The United States Centers for Disease Control and Prevention notes that no single assessment tool should be used as the sole basis for diagnosis; evaluation commonly draws on caregiver descriptions of development and professional observation.¹

A diagnosis may help organise relevant clinical knowledge.

It may identify co-occurring conditions.

It may support access to AAC, occupational therapy, educational accommodations, disability services, or other resources.

It may give adults a framework through which to reinterpret earlier experiences.

It may help families stop treating overload, communication difference, repetition, shutdown, or difficulty with transitions as deliberate misbehaviour.

It may protect someone from systems that otherwise demand proof of need every time support is requested.

Amanah should be able to store a diagnosis carefully, source it properly, restrict its disclosure, connect it to appropriate professional recommendations, and preserve its relevance where it matters.

But it should never allow the diagnosis to become the person’s governing identity.

The spectrum is not a single line

People often imagine the autism spectrum as a line.

At one end is “mild.”

At the other is “severe.”

A person is placed somewhere along it and expected to remain there.

That picture is too simple.

The DSM-5 criteria reproduced by the CDC specify support severity separately for social communication and for restricted or repetitive patterns of behaviour. They also separately specify whether autism is accompanied by intellectual or language impairment.¹

That alone should warn us against collapsing several different questions into one label.

A person may have:

• substantial communication support needs;
• relatively strong reasoning in some domains;
• significant sensory distress;
• low tolerance for unpredictable transitions;
• no spoken language;
• effective AAC in familiar settings;
• motor-planning difficulty;
• strong visual understanding;
• fluctuating speech;
• intellectual disability;
• no intellectual disability;
• significant daily-living support needs;
• or relative independence in selected areas.

These combinations are not interchangeable.

Neither are communication, cognition, adaptive functioning, sensory regulation, executive function, motor access, emotional regulation, and legal decision-making.

Amanah cannot encode:

autism → one fixed support package

It needs to understand the particular person.

A support level is not a biography

Clinical support specifiers can be useful shorthand.

They may help communicate that a person requires support, substantial support, or very substantial support.

But they do not tell Amanah:

• how this person says yes;
• how they say no;
• which sensory conditions cause pain;
• whether they use speech all the time, sometimes, or not at all;
• how long they need to process a question;
• which routines support regulation;
• what they understand;
• which forms of help they welcome;
• what they can do in a familiar environment;
• what becomes inaccessible under overload;
• which interests matter to them;
• who they trust;
• what they want to learn;
• or how their needs change across settings.

The support level may belong in the record.

It cannot substitute for the record.

The same principle applies beyond autism.

A diagnostic code may help classify a condition.

It does not contain the person’s life.

Speech status is not an intelligence test

Amanah began with a serious concern: non-speaking autistic people are often interpreted through the absence of conventional speech.

This can produce two opposite errors.

The first is pessimistic:

The person does not speak, therefore they do not understand.

The second is romantic:

The person does not speak, therefore there must be an intact or exceptional intelligence hidden underneath every difficulty.

Neither claim is safe.

The research report supporting Amanah found that terminology such as non-speaking, nonverbal, and minimally verbal is used inconsistently across studies. It also noted that inaccessible assessments may underestimate ability when they depend heavily on spoken instructions, speed, or motor responses.

The appropriate Amanah principle is:

Do not infer intellectual ability, decision capacity, preference, or communicative potential from speech status, response speed, apparent compliance, or a global support label.

That does not erase intellectual disability when it is present.

It prevents the system from making unsupported conclusions in either direction.

Amanah should preserve:

• accessible assessment;
• demonstrated understanding;
• communication access;
• professional findings;
• person expression;
• uncertainty;
• context;
• and revision over time.

A person should not have to perform competence through the evaluator’s preferred channel before the system treats them as present.

High masking and high support needs are not the same life

This programme grows partly from two very different vantage points.

One is the experience of a high-masking neurodivergent adult who has spent years navigating expectations, fitting in, becoming overwhelmed, losing access to abilities under pressure, and struggling to explain what is happening internally.

The other is the experience of caring for a non-speaking autistic child with high support needs whose communication and sensory world are more visibly incompatible with mainstream systems.

These lives should not be collapsed together.

High masking does not mean “mild.”

High support needs do not mean absence of personality, intelligence, pride, preference, or possible contribution.

They expose different failures in the same surrounding world.

The high-masking person may appear capable because they have learned scripts, suppressed visible distress, imitated expected behaviour, or paid for ordinary participation through exhaustion.

The high-support-needs person may be excluded because their difficulty cannot be hidden and the environment refuses to adapt.

Research on autistic camouflaging describes masking and compensation strategies used to fit in or connect socially, alongside reported costs including exhaustion and threats to self-perception. Camouflaging may also affect whether autistic characteristics are recognised during assessment.² Later research has continued to treat camouflaging as a potential barrier within adult diagnostic processes rather than a sign that support is unnecessary.³

Amanah must be capable of recognising support needs that are visible.

It must also be capable of respecting support needs that have been made invisible through adaptation.

Masking is not proof of independence

A person may hold a job and still require recovery after ordinary interaction.

They may speak fluently and lose speech under overload.

They may manage a complex intellectual project and be unable to make a phone call.

They may appear calm while experiencing severe sensory distress.

They may maintain eye contact because they learned that refusal is punished.

They may complete a task once and be unable to repeat it under different conditions.

They may organise one area of life meticulously while another collapses.

They may navigate public life through rehearsed rules that no observer can see.

Amanah should not treat successful performance as proof that support is no longer needed.

It should ask:

• Under which conditions was this possible?
• What preparation was required?
• What recovery followed?
• Which parts were scripted?
• Which supports were invisible?
• Does the person want to keep doing it this way?
• What happens when the environment changes?
• Is the apparent independence sustainable?

This is not a request to pathologise every achievement.

It is a refusal to make suffering invisible merely because the outcome looked acceptable.

Visible dependence is not total dependence

The reverse error is just as harmful.

A person may require support with:

• communication;
• personal care;
• transport;
• transitions;
• safety;
• food;
• sensory regulation;
• planning;
• or daily routines.

That does not mean they require somebody else to govern every part of life.

Dependence can be domain-specific.

A person may need extensive physical assistance while communicating clear preferences.

They may need a guardian for certain legal or financial decisions while directly choosing clothing, activities, relationships, meals, music, faith practice, or who may enter their room.

They may need a familiar supporter during medical appointments while still being the person whose assent, distress, refusal, and comfort matter.

They may contribute through art, domestic participation, memory, affection, focused skill, spiritual life, community presence, or work adapted to their actual access needs.

Amanah should not promise unsupported independence as the only dignified outcome.

It should support the greatest practicable agency, participation, communication, safety, and quality of life—with as much dependable assistance as the person actually needs.

Reliable interdependence is also a human life.

Diagnosis can explain without becoming destiny

A diagnosis may explain recurring patterns.

It may help the person or family anticipate support needs.

It may help professionals choose appropriate assessments.

But Amanah should never convert a diagnosis into a rule that says:

This person will always respond this way.

Autistic people change.

Children grow.

Communication develops.

AAC access improves or becomes less accessible.

Sensory needs shift.

Health changes.

Trauma changes what feels safe.

A person acquires new skills.

They reject old routines.

They become interested in something new.

They form relationships that alter how they navigate the world.

They may continue needing substantial support while the form of support changes.

A diagnosis may remain valid while its practical expression changes across the life course.

The continuity spine must therefore preserve both:

• enduring clinical context; and
• present person-specific reality.

The first should not erase the second.

The Care Profile should be functional, not diagnostic theatre

Amanah needs an individual support profile.

The profile may be informed by diagnosis, but it should be organised around what the person actually needs, prefers, communicates, experiences, and chooses.

A useful profile may include:

Communication

• current communication methods;
• AAC access;
• yes, no, stop, help, pain, and uncertainty signals;
• response time;
• motor or sensory access;
• communication that changes by setting;
• and interpretations that remain uncertain.

Sensory regulation

• sound, light, texture, smell, movement, touch, temperature, and proximity;
• early signs of overload;
• preferred regulation supports;
• supports the person rejects;
• recovery conditions;
• and the sensory footprint of the technology itself.

Executive and transition support

• task initiation;
• sequencing;
• preparation;
• visual schedules;
• warning intervals;
• pacing;
• demand load;
• recovery;
• and what becomes difficult during change.

Daily life and safety

• food and hydration;
• sleep;
• mobility;
• health references;
• current risks;
• emergency communication;
• personal care;
• transport;
• and environmental support.

Participation and possibility

• interests;
• demonstrated competencies;
• preferred tasks;
• accessible environments;
• accommodations;
• activities the person wants to explore;
• forms of contribution;
• refusals;
• and the conditions that made previous success possible.

Relationships and trust

• trusted people;
• support roles;
• privacy preferences;
• familiar rituals;
• relationship changes;
• and who the person does not want involved.

This is not an “autism profile.”

It is this person’s continuity.

Diagnosis must remain separate from personhood data

There is another technical reason to avoid making diagnosis the central object.

Diagnostic information is sensitive.

Not every participant needs to see it.

A teacher may need to know the approved communication and sensory supports without receiving the whole clinical record.

A transport worker may need travel accommodations without knowing the diagnosis.

A device may need a stop signal and safe distance without any diagnostic label at all.

An employer may receive selected accommodations only when the person chooses or the lawful process requires disclosure.

A robot carrying an AAC device does not need to know that the person is autistic.

It needs to know:

• the approved task;
• the destination;
• approach distance;
• current assent or refusal;
• and safe-state behaviour.

Separating diagnosis from functional support helps Amanah follow minimum-necessary disclosure.

It also protects the person from one label becoming visible everywhere the continuity travels.

No “autism mode”

One of the easiest design mistakes would be to create a permanent Autism Mode.

The interface changes.

The model adopts simplified language.

Every hesitation becomes sensory overload.

Every preference becomes rigidity.

Every repeated interest becomes pathology.

Every disagreement becomes a symptom.

Every grief response is interpreted through autism.

Every creative activity becomes therapy.

That would be unbearable.

A person is not autistic only when struggling.

They are not a diagnosis wrapped around a collection of interventions.

They may need help preparing for a noisy event today.

Tomorrow they may want to write a story.

Next week they may be grieving.

Later they may need support making a medical decision.

Then they may want to discuss a special interest for the pleasure of it.

Amanah’s Stem Cells should route relevant functions when the situation calls for them.

Sensory support when sensory context is relevant.

Communication support when communication access is relevant.

Transition support around transitions.

Clinical references in the proper clinical scope.

Ordinary companionship when no disability-specific intervention is needed.

The governing rule is:

Do not load the diagnosis over the whole person. Route the support required by the present situation.

The undiagnosed person still has needs

Formal diagnosis cannot be the only reason Amanah recognises support.

Some people are undiagnosed because they cannot afford assessment.

Some live far from qualified services.

Some encounter long waiting lists.

Some were children in systems with little autism awareness.

Some come from families or cultures where difference is minimised, stigmatised, spiritualised, hidden, or treated as a private shame.

Some adults learned to camouflage.

Some women and gender-diverse people were assessed through narrow expectations shaped by more historically recognised presentations.

Some people have needs that are real but do not fit neatly within one diagnosis.

Global autism research and diagnostic infrastructure are also uneven. A cross-cultural research framework published in Autism Research noted that autism research is heavily concentrated in Western high-income countries and that culturally appropriate screening and diagnostic tools are limited in many low- and middle-income settings.⁴

Amanah should not pretend it can diagnose these people.

It should not present self-identification, caregiver concern, a school observation, or an AI pattern as equivalent to clinical diagnosis.

But it can still distinguish between:

• diagnosis;
• provisional professional opinion;
• functional assessment;
• documented accessibility need;
• self-reported experience;
• caregiver observation;
• and machine hypothesis.

Low-risk support does not always require the system to decide why the need exists.

A visual schedule can be useful before diagnosis.

A sensory journal can be useful before diagnosis.

A communication passport can be useful before diagnosis.

A person-owned record of shutdown, speech loss, or environmental difficulty can be useful before diagnosis.

Amanah can help make needs legible without pretending to settle the clinical question.

The next article will address the harder access question directly:

Who gets through the gate when diagnosis is scarce, delayed, inaccessible, or culturally blocked?

Here, the principle is simpler:

Lack of diagnosis does not prove lack of need.

Diagnosis should not be generated by Amanah

Amanah may eventually hold observations that appear clinically relevant.

It may notice repeated sensory distress.

It may identify changes in sleep or movement.

It may detect that speech access fluctuates.

It may summarise long-term patterns for a professional appointment.

It may help a family reconstruct developmental history.

It may help an adult prepare questions for assessment.

But Amanah should not diagnose.

It should not say:

“You are autistic.”

It should not say:

“This child is Level 3.”

It should not infer intellectual disability.

It should not recommend a diagnostic label as though the model were a qualified clinician.

It should not turn a support profile into a hidden classifier.

It may say:

“These recorded experiences may be relevant to discuss with a qualified professional.”

Or:

“This summary is based on person-reported and caregiver-observed patterns. It is not a diagnosis.”

This boundary protects both the person and the legitimacy of the support system.

Amanah should be helpful before, during, and after assessment without impersonating the assessment.

A diagnosis should not automatically unlock every capability

Diagnosis should not become a master key either.

A formal diagnosis does not automatically mean:

• the person consents to continuous monitoring;
• a parent may see everything;
• a professional may control every function;
• a humanoid may enter the home;
• the system may record intimate care;
• a centre may share the person’s history;
• or an AI may make higher-risk decisions.

Capabilities should follow their own risk, evidence, permission, and regulatory requirements.

A diagnosis may help configure support.

It should not erase the Participation and Authority Gate.

This distinction matters because safety-gated technology can easily become diagnosis-gated technology.

The next two Amanah articles will develop that problem more fully:

• who is excluded when diagnosis becomes the universal doorway;
• and why permission requirements should follow the risk of a function rather than the prestige of a label.

Diagnosis can be liberating and still insufficient

For some adults, receiving a diagnosis offers relief.

It can provide a framework for years of difference, misinterpretation, or self-blame.

Systematic reviews of adults diagnosed later in life describe both the struggle to obtain assessment and the importance of post-diagnostic support; diagnosis may provide recognition without automatically repairing the practical difficulties that preceded it.⁵

That is important for Amanah.

A diagnosis can make a person more understandable to themselves and others.

But a label without accommodations, communication access, financial support, appropriate services, or a change in the environment may leave the person carrying the same life with a new word attached.

Amanah should not treat diagnosis as the completion of support.

It should ask what comes next.

What information does the person want preserved?

What support becomes available?

What changes at home, school, work, or in healthcare?

What remains inaccessible?

What does the person reject?

Which old interpretations need correction?

What has the diagnosis clarified—and what has it not clarified?

The diagnosis may begin a new continuity process.

It is not the continuity itself.

What the system should encode

Amanah should be able to represent diagnosis with precision.

For example:

Diagnostic record

• diagnosis or clinical formulation;
• diagnosing professional or service;
• jurisdiction;
• date;
• diagnostic system or criteria;
• relevant specifiers;
• source document;
• disclosure restrictions;
• review or reassessment information;
• and whether the person has seen or discussed the record.

Functional support record

• present communication;
• sensory needs;
• executive support;
• daily-living assistance;
• AAC;
• accommodations;
• safety;
• participation;
• and current person preferences.

Person-owned identity record

• language the person uses for themselves;
• identity terms they accept or reject;
• how they want the diagnosis discussed;
• who may know;
• and what they do not want reduced to diagnosis.

Professional recommendation

• specific recommendation;
• professional scope;
• intended context;
• evidence basis;
• review date;
• and whether it is active, advisory, superseded, or expired.

Unconfirmed need

• person report, caregiver observation, or functional assessment;
• no claim of diagnosis;
• support that may be offered safely;
• uncertainty;
• and possible referral pathway.

This structure allows diagnosis to matter without colonising the entire profile.

What Amanah must refuse

Amanah should refuse several common shortcuts.

It should refuse:

“Autistic people prefer routine.”

It should ask:

“Which routines matter to this person, and what happens when they change?”

It should refuse:

“Level 1 means independent.”

It should ask:

“Independent in which task, under which conditions, at what cost, and with which invisible supports?”

It should refuse:

“Level 3 means incapable.”

It should ask:

“What communication, decision, participation, and support pathways are accessible to this person?”

It should refuse:

“No diagnosis means no need.”

It should preserve functional need without manufacturing a diagnosis.

It should refuse:

“Diagnosis means permission.”

It should route every consequential capability through its proper gate.

It should refuse:

“Masking means coping successfully.”

It should consider sustainability, internal cost, and the person’s own account.

It should refuse:

“Visible dependence means no authority.”

It should preserve domain-specific participation.

It should refuse:

“The profile describes the person.”

It should remember that the profile is a revisable representation.

The governing sentence

A diagnosis can be clinically important.

It can be personally meaningful.

It can unlock resources.

It can correct years of misunderstanding.

It can guide professional support.

But it does not contain:

• the whole sensory world;
• the whole communication system;
• the whole intelligence;
• the whole support need;
• the whole identity;
• the whole legal capacity;
• the whole future;
• or the whole person.

Amanah must be designed around the individual life, not around a permanent diagnostic mode.

The diagnosis may enter the continuity spine.

It must never become the spine.

---

Research status

This essay extends the fourteen foundational Amanah Companions articles, particularly:

• What an AI Companion for a Non-Speaking Autistic Child Would Actually Need to Know
• Designing the Care Profile
• AAC Integration and the Communication Map
• The Sensory Map: Why Regulation Comes Before Compliance
• Human-Led Care in an AI-Assisted Future
• The Amanah Companion Framework: Spine, Stem Cells, and Care Memory

It also develops later authorial clarification concerning high-masking autistic or neurodivergent adults, non-speaking autistic people with high support needs, functional support profiles, diagnosis inequality, and context-sensitive routing rather than permanent “autism mode.”

This essay does not provide diagnostic advice. It proposes design principles for an assistive continuity system. Clinical diagnosis, reassessment, legal capacity, disability eligibility, and access to services remain jurisdiction- and profession-specific.

Sources

1. Centers for Disease Control and Prevention, Clinical Testing and Diagnosis for Autism Spectrum Disorder, updated 8 May 2025. The page notes that no single tool should be the sole basis for diagnosis, reproduces DSM-5 diagnostic criteria, separates severity across social-communication and restricted/repetitive-behaviour domains, and separately specifies intellectual and language impairment. https://www.cdc.gov/autism/hcp/diagnosis/index.html ↩
2. Laura Hull et al., “‘Putting on My Best Normal’: Social Camouflaging in Adults with Autism Spectrum Conditions,” Journal of Autism and Developmental Disorders 47 (2017): 2519–2534. https://doi.org/10.1007/s10803-017-3166-5 ↩
3. Alison U. Tassone et al., “Autistic and Clinician Perspectives on the Influence of Camouflaging in Adult Autism Assessments,” Research in Autism 131 (2026), article 202788. The study reports agreement among participating clinicians and autistic adults that camouflaging can influence assessment, while its sample and setting should not be treated as globally representative. https://doi.org/10.1016/j.reia.2026.202788 ↩
4. Rosa A. Hoekstra et al., “A Conceptual Framework for Understanding the Cultural and Contextual Factors on Autism Across the Globe,” Autism Research 13, no. 7 (2020): 1029–1050. https://pmc.ncbi.nlm.nih.gov/articles/PMC7614360/ ↩
5. Juwayriyah M. Nayyar et al., “Exploring Lived Experiences of Receiving a Diagnosis of Autism in Adulthood: A Systematic Review,” Autism in Adulthood (2025). The review synthesises reported adult experiences including difficulty obtaining diagnosis, stigma, haphazard post-diagnostic support, and the mixed personal meaning of receiving a diagnostic framework. https://pmc.ncbi.nlm.nih.gov/articles/PMC11937775/ ↩