A Lejacie framework · Chami Rupasinghe
Safe to Say, Safe to Reach, Safe to Use
An emerging framework for closing the Paper-to-Person Gap in regulated expertise so it can reach, matter to and safely help real people.
Start with the letterDownload the paper (PDF)
Cite this
Rupasinghe, C. (2026). Safe to Say, Safe to Reach, Safe to Use: An emerging framework for closing the Paper-to-Person Gap in regulated expertise so it can reach, matter to and safely help real people (Draft v0.1). Lejacie. https://lejacie.com/Safe-to-Say-Safe-to-Reach-Safe-to-Use-Lejacie.pdf
Safe to Say, Safe to Reach, Safe to Use (Draft v0.1, 2026).
Chami Rupasinghe, Safe to Say, Safe to Reach, Safe to Use: An Emerging Framework for Closing the Paper-to-Person Gap in Regulated Expertise so It Can Reach, Matter to and Safely Help Real People (Lejacie, Draft v0.1, 2026) <https://lejacie.com/Safe-to-Say-Safe-to-Reach-Safe-to-Use-Lejacie.pdf>.
We created Safe to Say, Safe to Reach, Safe to Use.
Safe to Say protects people from being misled. Safe to Reach protects people from being missed. Safe to Use protects people from being abandoned.
We studied exactly what makes this kind of information so hard for the human brain to use, so we could close the gap.
So professionals would not have to choose between “this has to be defensible, I don’t want to get sued, so I have to say these things” and a person actually understanding it, trusting it, relying on it and acting on it, in the moment they need to reach for it.
For the professional
Safe to say
Is it accurate, properly sourced, qualified, bounded and within the authority of the source and carrier?
For the person
Safe to reach
Can it reach the people who may need it, in time and through usable channels, with a truthful reason to recognise why it may matter?
For the person
Safe to use
Can they understand what matters, judge what they can rely on, take a feasible next step, seek help, question an error and recover when the first route fails?
Explore Safe to Say, Safe to Reach, Safe to UseDownload the paper (PDF)All ideas and methods
The Paper-to-Person Gap
Imagine receiving a letter.
It comes from a court, government department, insurer, employer, school, lawyer, hospital or regulator. It says something important has happened: a decision, deadline, right, risk, entitlement, obligation or process may affect you.
It says
You may be eligible.
Your circumstances may affect the outcome.
We cannot guarantee the result.
You should seek independent advice.
Time limits apply.
A person may not be able to:
- Find it before it is too late.
- See that it may matter to them.
- Recognise that a right, risk, choice, deadline or route to help is involved.
- Understand what is happening in practical terms.
- Work out what they can rely on and what they should not assume.
- Identify who decides and what remains uncertain.
- See the questions they need answered before they know to ask them.
- Take a safe, realistic next step.
- Follow through when time, stress, competing responsibilities or a practical barrier intervenes.
- Reach help.
- Correct, challenge, review, complain, escalate or recover if the first route fails.
The letter may be completely correct. That is the problem.
The problem
Correct at its source. Still failing the person it was meant to help.
The gap opens when expertise is organised according to the logic of the profession, while the person is left to turn it into the logic of their own situation, judgement and action. The problem is not that people are careless or incapable. The problem is that institutions often treat publication, legal sign-off, a referral or a document being sent as evidence that expertise has transferred. It has not.
- It cannot be found, accessed or reached in time“I did not know this existed until it was too late.”
- It gives no reason to attend“I did not realise this mattered until it was too late.”
- It feels generic or institutional“I do not know whether this is about me.”
- It gives facts without a practical picture“I read it, but I still do not know what is happening.”
- It assumes knowledge the person does not have“I did not know I needed to ask that, bring that document or take that step.”
- It hides the limits of the answer“I do not know what I can rely on.”
- It gives a next step that is not feasible“I meant to deal with it, but I could not make it happen.”
- It ends with a disclaimer or referral“I still do not know what to do, or where to go if that fails.”
- It offers no way to correct or challenge an error“The system is wrong, but I do not know how to question it.”
The failure is a failure of transfer.
Institutions often use completion of an internal task as evidence that the external task has been completed.
| Institutional completion | What may still be unresolved for the person |
|---|---|
| The policy is approved | “How does this affect me?” |
| The legal content is correct | “What can I actually rely on?” |
| The notice has been sent | “Did I see it, understand it and receive it in time?” |
| The webpage is published | “Can I find it when I need it?” |
| The material is in plain language | “What should I do in my circumstances?” |
| The form is online | “Can I complete it on my phone, with my documents and in private?” |
| The referral has been given | “Will that service help me, and what happens if it cannot?” |
| The decision includes review rights | “Can I realistically ask for review before the deadline?” |
| The AI system gives an answer | “Is this general information, a recommendation or something I can safely rely on?” |
| The service process exists | “Can I get through it while managing work, care, money, safety, disability, language or stress?” |
SourceCarrierPersonJudgementActionHelp, review or recovery
The framework shifts the unit of analysis from the document, message, tool or isolated interaction to the full pathway through which expertise must travel. At every point, something can be lost, distorted, overclaimed, hidden, delayed, made inaccessible or left without a workable route forward.
What must travel
Source
A person needs to know where the information, rule, finding or decision comes from and whether it is official, current and applicable.
Authority
A person needs to know who is entitled to explain, advise, assess, decide, review or make a binding determination.
Meaning
A person needs a practical account of what is happening, rather than a conclusion alone.
Conditions
A person needs to see the criteria, facts, exceptions and dependencies that may affect whether the expertise applies.
Uncertainty
A person needs to know what is not known, what may change and what cannot be concluded from the available information.
Relevance
A person needs a truthful reason to recognise why the expertise may matter to them.
Hidden questions
A person needs help identifying the facts, documents, choices or questions that an expert would know to consider.
Boundary
A person needs to distinguish general information from personalised advice, assessment, prediction or decision-making.
Action
A person needs a next step that is specific enough to be feasible in their circumstances.
Handover
A person needs to know who can take over when individual judgement, representation, assessment or decision-making is required.
Recovery
A person needs a route to correct an error, seek review, complain, escalate, re-enter the pathway or obtain alternative help.
What would have to be true for a person, in realistic conditions, to use this expertise safely, and who is responsible for making those conditions possible?
The conditions are not exceptional
Build for the conditions in which people actually need help.
In the public-serving systems where people seek legal, health, financial, social and regulatory help, the conditions that make expertise hard to use are often not unusual cases at the edge of the system. They are the ordinary conditions in which people arrive.
Source: Victoria Legal Aid, Community Legal Services Program report 2024–25.
Do not build for the calm, confident, well-resourced person with time, privacy, complete documents, stable internet and ready access to advice, then call everyone else a special case.
The professional dilemma
Professionals often have to help a person without having all the facts, holding the authority to decide the outcome or being entitled to give individual advice. The professional must not go too far, while the person cannot be left with too little.
Without a method, professionals often protect the boundary through caveats, disclaimers and referrals. Those may be necessary. But if they are not paired with practical meaning, a real handover and a path forward, caution becomes withdrawal.
Safe to Say, Safe to Reach, Safe to Use creates a third option: preserve the boundary, explain everything that can safely be explained and build the pathway at the point where expertise must be handed over.
What the gap costs
| For people | For professionals | For institutions and systems |
|---|---|---|
| Missed rights, deadlines, reviews, payments, remedies or services | Repeat explanation, triage, repair work and moral distress | Avoidable complaints, delay, escalation and unequal outcomes |
| Unsafe or delayed action | Time spent compensating for unclear resources and broken handovers | Resources that are technically complete but fail in practice |
| Over-reliance on incomplete or unsuitable information | Pressure to give individual answers without authority or enough facts | Records that mistake system failure for “did not proceed” or “failed to respond” |
| Reliance on confident but unreliable AI, scam, influencer or misinformation sources | Defensive retreat into caveats, referrals and disclaimers | Loss of warranted trust and a growing advantage for false certainty |
| Carrying the work of navigating between services, explaining the same story and recovering from failed pathways | Burnout caused by repeatedly repairing preventable transfer failures | Reforms that exist on paper but do not change lived reality |
The standard
Three conditions. They must hold together.
Expertise that is easy to find but inaccurate is unsafe. Expertise that is accurate but impossible to locate, understand or act on is insufficient. Expertise that is clear and engaging but exceeds the authority of its source or carrier is unsafe.
Try it. Switch each condition on and off.
Safe to Say protects people from being misled.
Safe to Reach protects people from being missed.
Safe to Use protects people from being abandoned.
What safe means
Safe does not mean risk-free, guaranteed to produce the correct outcome or a substitute for individual advice, professional judgement or a formal decision.
Safe means designed to support understanding, warranted reliance and feasible action while making material limits, uncertainty, authority and escalation needs visible.
Shared use
A source owner may be responsible for accuracy and authority; a communications or community team may lead reach; a service, design or frontline team may lead usability; and a governance or quality function may oversee the whole pathway. No single team has to perform every task. However, someone must be accountable for whether the three conditions hold together.
The three conditions in detail
Can we stand behind it?
Is this expertise accurate, properly sourced, qualified, bounded and within the authority of the source and carrier?
Safe to Say is the compliance, professional responsibility and authority part of the standard. It means that what reaches the person is still something the relevant source owner, professional or institution can stand behind. It means preserving what carries meaning while changing the form freely where the meaning survives.
| Question | What it checks |
|---|---|
| What is the authoritative source? | Law, policy, evidence, professional standard, formal decision or identified judgement |
| What exactly does it establish? | The claim that can safely be carried |
| What could change the answer? | Conditions, facts, thresholds, exceptions, missing information and time limits |
| What remains uncertain? | Evidence limits, disagreement, unknown facts or a decision not yet made |
| Who has authority? | Who can explain, advise, assess, decide, review or formally determine |
| How far can this carrier go? | What a lawyer, frontline worker, website, caseworker, educator or AI system may properly say |
| What must the resource not imply? | Eligibility, certainty, predicted outcome, individual advice or formal determination |
Do not say more than the expertise can support. Do not say so little that the person is misled by omission or left without a usable path. A disclaimer is not a boundary plan. The boundary must be designed into the pathway.
If it is absent: People may be misled, over-rely on content, receive unauthorised advice or act on claims that cannot be supported.
Safe to Say protects people from being misled.
Will it reach the people who need it and help them recognise why it matters?
Can the expertise reach the people who may need it, in time and through usable channels, with a truthful reason to recognise why it may matter?
Safe to Reach is the attention, relevance and human-cognition part of the standard. A person cannot use expertise they do not notice, cannot locate, do not recognise as relevant or do not care enough to stay with. Safe to Reach does not turn regulated expertise into marketing. It puts attention and engagement under a stricter ethical rule: the care it creates must be care the facts genuinely warrant.
| Question | What it checks |
|---|---|
| Will the right person encounter this? | Findability, delivery route, timing and access |
| Can they use the channel? | Phone, digital access, language, privacy, disability access, safety, cost and time |
| Will they recognise themselves? | Relevance cues, examples, categories and situations |
| Can they see the real stake? | Rights, risks, consequences, choices, opportunity or deadline |
| Is the urgency real? | Proportionate timing without manufactured panic |
| Does it acknowledge the human reality? | Fear, shame, grief, anger, distrust, overload or anticipated consequences without exploiting them |
| Does it have an accountable voice? | Who is speaking, why they know and who is responsible |
| Does it give a reason to continue? | A visible next move, not a vague instruction or institutional preamble |
The moment the feeling outruns the facts, Safe to Reach has failed Safe to Say. Safe to Reach must never use false urgency; fear, shame or humiliation; scapegoating, grievance or identity pressure; hidden persuasion; false certainty; selective omission of material information; a personal story that implies an unwarranted outcome; apparent personalisation that exceeds the carrier’s authority; or an action prompt designed to secure compliance rather than support informed judgement.
If it is absent: Expertise remains technically correct but unseen, ignored, received too late or encountered without sufficient relevance or trust.
Safe to Reach protects people from being missed.
Can they use it when it matters?
Has enough meaning, context, limitation, pathway and recovery route travelled with the expertise for a person to make a safe judgement and take a feasible next step?
Safe to Use is the legal-design, human-centred-design and real-life pathway part of the standard. It addresses the problem experts often cannot see unaided: a person does not know what they do not know to ask. It is designed for the conditions in which people often need help, not for an ideal reader with time, money, confidence, privacy, complete documents, stable internet and ready access to advice.
| The person’s question | What the expertise must provide |
|---|---|
| “What is happening?” | A clear account of the situation, roles, sequence, decision and practical consequence |
| “What else might matter?” | The questions, documents, deadlines, conditions and decision points the person may not know to ask about |
| “What can I rely on?” | What is established, conditional, uncertain, incomplete, outside scope or for someone else to decide |
| “Who decides?” | The difference between a source, explainer, adviser, assessor, decision-maker and reviewer |
| “What can I do now?” | Rights, choices, risks, consequences and a feasible next step |
| “Can I actually manage this?” | A practical plan: when, where and how to act; what to have ready; likely obstacles and a backup |
| “Who can help me?” | A named service, contact route, access details, hours, eligibility and what to say |
| “What if this is wrong or does not work?” | Review, correction, complaint, appeal, escalation, independent help, fallback and recovery routes |
Where the person cannot act because of safety, money, time, capacity, access, service failure or absence of a meaningful choice, the method identifies a system finding. It does not treat the person as the failure.
If it is absent: People may understand fragments but remain unable to act, seek qualified help, question an error or recover when the intended route fails.
Safe to Use protects people from being abandoned.
Governing principles
Ten non-negotiable commitments.
They guide decisions where the ten steps, delivery channels or local implementation details vary, and keep the framework grounded in authority, truthfulness, practical usefulness, fairness and accountability.
01Preserve source integrity
Expertise must remain faithful to what its source is entitled to support. Do not simplify by removing material conditions, uncertainty, exceptions, time limits, evidentiary limits or qualifications. Do not present a general rule as a personal conclusion, or an emerging finding as settled fact.
Test: Would the source owner recognise this as an accurate and properly bounded representation of the underlying law, evidence, decision, policy or professional judgement?
02Match claim to authority
No carrier should say, imply, recommend, assess or decide more than their role, competence and authority permit. The framework does not require every person, website or AI system to provide individual advice. It requires each carrier to make its boundary visible, explain what it can safely explain and offer a real pathway to authorised help where individual judgement is required.
Test: What is this carrier entitled to do, and what must it hand over?
03Make relevance truthful
People need a clear reason to attend to consequential expertise. That reason must be warranted by the facts, evidence, risk and person’s likely circumstances. The framework permits direct warnings, serious language and genuine urgency where they are justified. It rejects manufactured fear, urgency, certainty, pressure or emotional manipulation as substitutes for relevance.
Test: Does the communication explain why this may matter without overstating what is known, likely or required?
04Design for real conditions
Expertise must work under the conditions in which people actually encounter it, not only in ideal conditions imagined by the institution. The framework treats these conditions as design inputs, not user shortcomings.
Test: Could a person use this safely at 4:47pm, under ordinary constraints and after the first route has failed?
05Support judgement, not compliance
The purpose is to enable informed, safe judgement and action, not to secure attention, completion or behavioural compliance at any cost. Where the person has a genuine choice, the framework should preserve that choice.
Test: Does this support the person’s understanding and agency, or does it merely drive a preferred institutional outcome?
06Make boundaries useful
A professional, institutional or technical boundary should not become a dead end. A boundary statement should be accompanied, where possible, by an explanation of what the resource can do, what further help is needed, how to reach it and what to do if the first route fails.
Test: Does the boundary protect against inappropriate reliance while leaving the person with a usable route forward?
07Treat access as pathway access
Access is more than being able to open a document or website. A resource is not fully accessible if the route it describes is inaccessible in practice.
Test: Can intended users access and use the whole pathway, not merely the first resource?
08Build in contestability and recovery
People must be able to question, correct, review, complain, escalate or seek alternative help when information, decisions, services or digital systems fail. A pathway with no realistic recovery route shifts too much risk onto the person.
Test: If this is wrong, unavailable, misunderstood or unsuccessfully acted on, what can the person do next?
09Share accountability across the pathway
No single team necessarily controls the whole journey, but the institution must not allow distributed responsibility to become unowned failure. The person experiences one system, not separate internal functions.
Test: Who owns accuracy, reach, accessibility, action, handover, correction and recovery, and can those owners work together?
10Apply proportionately and learn
The level of review, testing, documentation and governance should reflect the consequence of getting the transfer wrong. In every case, the framework should be tested, monitored and revised in response to evidence of failure.
Test: Is the level of care proportionate to the likely harm of misunderstanding, inaction, delay, over-reliance or pathway failure?
The method
First know what may travel. Then make it reach the person. Then make it usable. Then check that nothing important was lost.
The method does not begin by rewriting a document. It begins by asking: what has to travel with this expertise for a real person to use it safely? Choose a step.
Reach and Use may never outrun what Say can defend.
The minimum safe transfer
Enough must travel for the person to avoid foreseeable misunderstanding, inappropriate reliance or avoidable abandonment.
- What this is and where it comes from.
- Why it may matter, and for whom.
- What it establishes, and what it does not establish.
- Any critical conditions, deadlines, uncertainty or limits.
- What a person can reasonably do next.
- What requires individual advice, assessment, decision-making or another authorised form of help.
- How to reach that help.
- What to do if the first route fails or the information appears wrong.
The appropriate level of detail depends on the consequences of error.
The stress test
“4:47pm” names a condition, not a literal testing time: limited capacity, incomplete information, competing demands and a narrowing window for help. Tuesday keeps the situation ordinary and recognisable.
- Tired, frightened, angry, grieving or overloaded.
- Some facts are missing.
- A deadline is close.
- Another service closes at five.
- A child needs collecting.
- The phone is nearly flat.
- English is not their first language.
- No private place to make a call.
- Not sure they can trust the source.
- The professional helping has another person waiting.
If it works only for an ideal reader, the Paper-to-Person Gap remains open.
Run the test on your own resource
Think of one notice, factsheet, referral or AI answer. Under those conditions:
The professional test
If the resource works only because an exceptional frontline worker repeatedly fills its gaps, the resource itself is incomplete.
The system test
A two-hour wait on the published number. A form that does not work on a phone. A named service with no capacity. No interpreter pathway. These are not failures of understanding, motivation or engagement. They are system findings.
Could the person use this when it mattered, with its meaning, limits and authority intact, and a feasible route forward?
The carrier matters
The carrier’s authority sets the ceiling for how far the expertise may be personalised.
| Carrier | What they may commonly do | What may require a boundary or handover |
|---|---|---|
| Authorised professional | Explain, apply expertise to facts, advise within their scope and make recommendations where authorised | Formal decisions or specialised advice outside their role |
| Court, regulator or decision-maker | Issue a formal determination, explain official process and identify review rights | Independent advice, advocacy or safety planning |
| Frontline worker or community practitioner | Explain general process, name public pathways, provide generic examples, support access and make referrals | Applying rules to individual facts, predicting outcomes or recommending a personal course |
| Public website, form or guide | Explain general rules, conditions, pathways, deadlines and support routes | Assessing an individual position or determining what they should do |
| AI-supported system | Retrieve approved information, explain general material, guide users through public pathways and escalate | Personalised legal, clinical, financial or other regulated advice unless specifically designed, authorised and governed to do so |
| Peer worker, advocate or support person | Share lived experience, explain practical pathways, help a person prepare questions and support access | Making professional determinations or presenting personal experience as authoritative advice |
The framework does not treat this as a reason to stop being useful. It asks: what can this carrier safely carry, and what must be handed over to someone with the authority to take the next part of the journey?
Worked example
From a police-issued safety notice to a pathway that exposes what still fails.
Illustrative notice created for this example. This fictional composite is not based on an actual court, police or service form. It is not legal advice, safety advice or a substitute for individual risk assessment, safety planning, legal advice or emergency help. If a person is in immediate danger, they should call 000.
Family Violence Safety Notice
Police have applied to the Magistrates’ Court for a family violence intervention order.
Your court date is [date] at [location].
Conditions set out in this notice apply to the respondent from [time/date] until the court deals with the matter or the notice otherwise ends.
You may be required to attend court.
If you need help, contact the court or seek legal advice.
Maya reads the notice late at night after the children are asleep. She and her former partner share a phone plan, and she worries he can see messages, call records or browser history.
The notice may be legally accurate. But it still leaves Maya carrying work that the source, pathway and surrounding expertise could have carried: locating the conditions, identifying who is bound by them, understanding what the court date means, working out which service can answer which question, and recovering when the available contact route is unsafe or inaccessible.
That is the Paper-to-Person Gap.
The rebuilt opening
You have received a family violence safety notice.
Police have applied to the court for a family violence intervention order.
The conditions listed under [heading] on page [x] currently apply to the respondent from [time/date].
Your court date is [date] at [time].
This page explains what the notice means, what the court will decide and where you can get support. It also explains what to do if the contact details in the notice are not safe or possible for you to use.
Where the pathway still fails
Maya cannot safely call from home. She tries the court’s online form using a borrowed device. The form requires an email address that she cannot safely use. The notice provides no other verified route. The framework does not solve that by writing “Use another safe route.”
| What Maya can now establish | What still needs qualified help | What remains a system failure |
|---|---|---|
| What notice she has received | What is safest for her personally | No verified alternative court contact route |
| Whether conditions are currently recorded as applying to the respondent | What legal option she should take | A form requiring an unsafe email address |
| Where to find the conditions | Whether she should attend or what the court is likely to decide | No confirmed fallback when standard channels cannot be used |
| What the court date is for | How a particular service will handle her details and circumstances | The responsible court or service must design and publish a workable alternative |
It does not make the missing route disappear. It makes the missing route visible, gives it an owner and stops the system from recording Maya’s inability to use it as a failure to engage.
Scope, limits and evaluation
A proposed framework, not a validated theory, professional standard or assurance scheme.
The framework is not a substitute for individual legal, clinical, financial, migration or other specialist advice; formal decision-making; risk assessment or safety planning for a specific person; regulatory compliance advice; audit, assurance or certification; a full accessibility assessment; a complete service-design process; a full AI governance system; or reform of an unjust rule, inadequate service or under-resourced system.
It can identify where these are needed. But it cannot make a missing service exist, make an unfair rule fair, guarantee a person’s safety or create meaningful choice where none exists. Where those conditions prevent use, the framework treats them as system findings, not failures of the person.
The framework has not yet been empirically validated. Its value will depend not on the appeal of its language, but on whether it helps institutions identify preventable transfer failures, assign responsibility for them and improve the real conditions in which people must understand, judge and act on consequential expertise.
A small, well-observed pilot is more valuable than a large rewrite programme that is never tested in practice.
Research fields informing the framework
It does not replace these fields. It connects selected responsibilities from them.
| Research field | What the framework draws from it |
|---|---|
| Legal design and people-centred justice | Participation, human-centred legal resources, justice journeys, practical access and designing legal systems around people’s needs and experiences |
| Access to justice, legal capability and expertise | Problem recognition, legal confidence, access to suitable help and the reasoning experts supply that non-experts may not possess |
| Plain language, information design and cognition | Findability, comprehension, structure, cognitive demands, working-memory limits and the limits of simplifying consequential expertise |
| Information behaviour and everyday information seeking | Recognising information needs, searching without expert vocabulary, sense-making, uncertainty, incidental encounter and use of informal intermediaries |
| Human-centred and service design | Context research, problem definition, prototyping, user testing, journey mapping and examining a resource within its surrounding service pathway |
| Knowledge translation and implementation science | Contextual adaptation, barriers and enablers, implementation feasibility, ownership, evaluation, monitoring and sustainment |
| Administrative burden | Learning, compliance and psychological costs imposed by institutional processes; distinguishing communication problems from structural and operational barriers |
| Health literacy and checking understanding | Access, understanding, appraisal and use; teach-back; and checking whether meaning has transferred rather than assuming that information delivery succeeded |
| Risk and uncertainty communication | Balanced risk information, probability, uncertainty, evidence limits, calibrated claims and warranted trust |
| Community engagement and public-interest communication | Listening, trusted relationships, relevant channels, community feedback, proportionate urgency and communication that supports informed protective action |
| Behaviour change and intervention evidence | Capability, opportunity and motivation; barriers to follow-through; and caution about assuming that behavioural interventions reliably produce intended effects |
| Shared decision-making and decision support | Balanced options, benefits, harms, uncertainty and personal values where a meaningful choice exists |
| Procedural and administrative justice | Voice, respectful treatment, intelligible reasons, authority, institutional decision-making, contestability and meaningful correction or review |
| Human factors, safety and handovers | Constrained conditions, workload, ordinary variation, transfer of responsibility, receiver acknowledgement, failed routes, adaptation and recovery |
| Trauma-informed practice and cultural safety | Safety, control, disclosure, power, historical experience, cultural context and avoiding interactions that cause further harm |
| Epistemic justice and co-production | Whose knowledge counts, credibility, lived expertise, participation, institutional blind spots and the burdens or risks of co-production |
| Accessibility, language access and digital inclusion | Accessible formats, disability access, assistive technology, interpreting, multilingual communication, channel choice, device constraints and alternatives to digital pathways |
| Professional responsibility and information governance | Authority, competence, confidentiality, scope, source provenance, version control, currency, records, correction and accountability |
| AI governance and appropriate reliance | Reliability, unsupported claims, human oversight, actual reliance, verification, escalation, monitoring and contestability |
The proposed distinction lies in the organisation of responsibilities, rather than in the invention of its component ideas. Preserve what expertise can responsibly support; enable it to reach the people who may need it; and make its meaning, limits and routes forward usable in real life.
Key points
Correct at the source is not enough.
The relevant unit of analysis is not only the document, message or tool. It is the full pathway: source, carrier, person, judgement, action, help, review or recovery.
The three conditions must hold together.
Clear but unauthorised content is unsafe. Accurate but unreachable content is insufficient. Engaging content without a feasible pathway can create attention without support.
A referral is not a complete handover.
A referral becomes a more credible handover when the person can understand where they are going, why, what will happen next and what to do if the service cannot help.
Failure often belongs to the system.
The framework asks institutions to make those points visible, assign them to an accountable owner and distinguish preventable transfer failures from limits that require broader reform.
Expertise is not safely transferred merely because it has been approved, published, sent, explained, made accessible or referred onward. It must also remain safe to say, safe to reach and safe to use.
Download the paper (PDF)The Paper-to-Person Gap
Cite this
Rupasinghe, C. (2026). Safe to Say, Safe to Reach, Safe to Use: An emerging framework for closing the Paper-to-Person Gap in regulated expertise so it can reach, matter to and safely help real people (Draft v0.1). Lejacie. https://lejacie.com/Safe-to-Say-Safe-to-Reach-Safe-to-Use-Lejacie.pdf
Safe to Say, Safe to Reach, Safe to Use (Draft v0.1, 2026).
Chami Rupasinghe, Safe to Say, Safe to Reach, Safe to Use: An Emerging Framework for Closing the Paper-to-Person Gap in Regulated Expertise so It Can Reach, Matter to and Safely Help Real People (Lejacie, Draft v0.1, 2026) <https://lejacie.com/Safe-to-Say-Safe-to-Reach-Safe-to-Use-Lejacie.pdf>.