Workers compensation

For a non-catastrophic injury, subsection 29(5) excludes compensation for any week within the 28 days beginning on the date of the injury, unless the relevant authority determines otherwise on the ground of financial hardship or the need to provide for adequate supervision of dependent children. Section 29A applies to catastrophic injury and has no 28-day exclusion and no weekly cap. An authorised person confirms which provision applies. From AI Can Map a Section 29 Household Services Claim. It Cannot Decide What Is Reasonable

Yes. Comcare's reconsideration guidance says a request need not be written, can be implied by the circumstances and should be clarified when correspondence is unclear. A model must therefore flag a possible request rather than dismiss one because the person wrote "complaint", and it must never declare that a request exists without human legal and procedural review. From A Complaint Is Not a Reconsideration: AI Can Route the Issue, Not Close It

No. Section 23A applies where the Commonwealth made salary, wages or pay and Comcare determined the compensation, with section 5(7) extending Commonwealth references to Commonwealth authorities. It deals with repayment, set-off and leave-credit restoration in that setting. A licensed corporation must check the provision actually engaged, the Fair Work Act, the applicable instrument and its own approved procedures. From Backdated Incapacity Needs Three Ledgers, Not One Payroll Fix

No. Regulation 11A(3) contains five defined exclusion situations, each with a prescribed start and endpoint, and Comcare confirms the section 71 employer-information notice exclusion applies only to Comcare. A licensee's request to its own payroll, HR or a manager does not exclude time, and an informal email is not automatically a section 58 notice. From Section 54 Intake: AI Can Find Gaps, Not Move the Clock

No. Comcare's scheme guidance states that the rehabilitation authority's section 36 power does not extend to assessing an employee's need for household services. Relevant authorities have separate powers to arrange and pay for a needs assessment under section 70 for Comcare or section 108F for licensees. A rehabilitation assessment must be taken into consideration but cannot be the sole basis for the household services decision. From AI Can Map a Section 29 Household Services Claim. It Cannot Decide What Is Reasonable

No. Comcare's pre-liability treatment reimbursement and rehabilitation cost arrangements sit inside a section expressly limited to Australian Government agencies and statutory authorities. A licensee's authorised people must apply the executed licence, the SRC Act, internal funding policy and applicable employment arrangements instead. From Early Support Before Liability: Track the Process, Not the Person

No. Section 5A defines injury with its employment connection and the reasonable administrative action exclusion, section 5B defines disease by employment contribution to a significant degree, and section 14 states the core compensation liability. A Nature or Mechanism code classifies harm and occurrence for data purposes. It proves none of those statutory elements. From TOOCS Coding Is Data Quality, Not a Liability Finding

No. Assessing whole person impairment against the approved Guide is a clinical judgement reserved for a suitably qualified medical practitioner, and the determination is a statutory decision for an authorised delegate. AI can organise the evidence, but it cannot decide the percentage or the entitlement and has no standing to. From AI and Permanent Impairment: Organise the Evidence, Keep the Judgement

No. Under section 72 of the SRC Act, Comcare is guided by equity, good conscience and the substantial merits and is not bound by the rules of evidence, and the statutory and delegation framework assigns determinations to authorised decision-makers. Tone analysis, sentiment labels and credibility percentages have no place in the ledger; stress, language differences or disability must never become machine proxies for credibility. From Statement Summaries Need Source IDs, Not a Neat Story

AI can check transparent arithmetic using verified inputs, but it must not determine the reasonable amount or apply an unverified statutory maximum. The subsection 29(1) weekly maximum is indexed each 1 July and published by Comcare, and the amount also cannot be less than 50 per cent of what the employee pays for the services. The authorised claims manager confirms the current rate, the evidence and the final calculation. From AI Can Map a Section 29 Household Services Claim. It Cannot Decide What Is Reasonable

No. AI can triage a de-identified file to flag that an overlap may exist and build a chronology of the parallel claim, which is useful early screening. The offset or recovery amount is a calculation that turns on the exact heads of damage, the sections engaged and the amounts involved, and it is a determination the case manager makes. An AI-suggested figure should never be relied on as the number. From Preventing Double Payment Under the SRC Act: AI Can Flag the Overlap, Not Calculate the Offset

No. In this design the skill never states whether liability is accepted or rejected and never drafts or redrafts the outcome sentence. The determination, its reasoning and every word of its outcome remain the delegate's. The skill is limited to checking that the evidence chain behind the draft is complete. From A Determination Evidence-Check Skill File Under Section 14

No. The control principle is that AI may organise information, but people must decide. Suitable duties need current medical evidence, knowledge of real work demands, consultation with the injured employee and review when circumstances change. Authorised people make the decision, and AI cannot replace that judgement. From AI Can Organise Recovery-at-Work Information, but People Must Decide

No. AI can build the analysis on a de-identified file: a chronology, the facts sorted against the subjective and objective limbs, an unable-versus- unwilling read, and a list of missing evidence. The reasonable-excuse finding is an exercise of delegated judgement on the specific facts, and any suspension, refusal or reinstatement that follows is the delegate's decision, recorded in their own reasons. From Reasonable Excuse Under the SRC Act: AI Can Structure the Analysis, Not Make the Finding

No. Section 5A is a chain of human judgements: characterising the condition, weighing employment contribution, classifying each action, and forming a view on whether conduct was reasonable and reasonably done. Those are evaluative calls a delegate is accountable for under the Act. A model that appears to decide them is worse than useless, because it produces a confident answer with no accountable reasoning behind it. From AI and the Reasonable Administrative Action Exclusion: Map the Actions, Keep the Judgement

No. Section 16(1) conditions compensation on the treatment being reasonable for the employee to obtain, and leaves the amount to what the authority determines is appropriate. In practice a properly delegated officer decides both. AI can assemble the picture: the category mapping, the evidence, the cost against the benefit and the gaps. The evaluative weighing and the determination itself stay with the human decision-maker, and a determination made without the delegation is invalid no matter how good the work-up was. From AI Can Build the Section 16 Picture, Not Make the Call

No. AI can prepare and organise the material and draft the referral, but the decision to require an examination under section 57 is a determination the relevant authority must make, and it must comply with the Guide. A model cannot weigh the necessity of the examination, the employee's personal circumstances or the choice of practitioner, and it cannot own a decision that carries a suspension consequence and a right of review. The delegate decides and signs, and a human reviews every AI output first. From Section 57 Examinations: AI Can Build the Referral, Not Make the Call

No. Comcare guidance is clear that medical professionals do not determine liability, and a model has even less standing to. AI can organise and compare conflicting opinions so the differences are visible, but the judgement about which opinion to prefer, and the liability determination under section 14 of the SRC Act, stays with the delegate, made on the balance of probabilities. From Two Doctors Disagree: AI Can Map the Conflict, Not Resolve It

No. The determination under the SRC Act is made by the determining authority through an authorised delegate, in practice usually the case manager, and a section 62 reconsideration must be conducted by a person who was not involved in making the original determination. AI output supports those decisions; it does not make them. The human stays accountable for the regulated act, and that accountability must be documented in the workflow design, not just in policy. From AI Tools in Workers Compensation Claims: Where Value, Where Risk, Where Governance

No. Subsection 36(2) requires the assessment to be made by a legally qualified medical practitioner nominated by the rehabilitation authority, a suitably qualified person other than a medical practitioner nominated by the authority, or a panel of such people. AI can organise de-identified material for the workflow, but it is not the statutory assessor and cannot determine the employee's capability of undertaking a rehabilitation program. From AI Can Organise a Section 36 Rehabilitation Assessment. It Cannot Choose the Program

No. The SRC Act 1988 was written for delegated human decision makers, and decisions under sections 14, 16, 19 and 24 can only be made by an appropriately delegated person. AI can prepare a draft of any of these, but it cannot lawfully make the legal decision itself. From SRC Act and AI Assisted Determinations: A Practitioner Framework

No. AI can freeze the population, apply the published band and expose coverage gaps, but the tool expects the auditor to choose an appropriate sample and to explain any alternate sampling methodology in the audit report. An AI risk ranking deliberately overweights unusual files, so it cannot stand in for a representative selection method. From A Defensible Claims Audit Sample Starts With the Population, Not an AI Risk Score

Yes. Comcare says early intervention may be used for injuries and illnesses that are work related or non work related, can happen whether or not an employee has lodged a workers compensation claim, and should be voluntary for an employee. It also says early intervention does not affect an employee's ability to make a claim. From Early Support Before Liability: Track the Process, Not the Person

No. Claimant material is sensitive health information and the matter ends in a statutory decision under the SRC Act. The safer pattern is to de-identify everything first, then use an LLM only to build an offline tool tested with synthetic rows, so real claimant data never reaches an external service. From Build a WC Evidence Chronology Tool Without Outsourcing Judgement

Yes, to prepare, never to conduct. AI can draft MI-informed openings, open questions, example reflections and follow-up messages so the human arrives rehearsed rather than improvising. The script is a rehearsal aid, not a teleprompter. Reflective listening cannot be pre-drafted, only rehearsed, and every draft needs human review first. From AI Can Draft Recovery Conversation Scripts, but the Listening Stays Human

Yes, for the reading task. AI structures long reports into diagnosis, history, treatment, prognosis, and recommendations, surfaces inconsistencies across reports, and reduces cognitive load on entry tasks. It is a navigation aid only. The first read is end to end, every time, and the summary never replaces the source report. From Treating Practitioner Reports and AI: Where the Workflow Helps and Where It Hurts

No. Where delay is reasonable, the manual says Comcare may, on the merits, temporarily extend a treatment plan using other current medical information, support the employee to obtain a certificate and maintain communication. That is not a general rule continuing incapacity liability or payments, and it is not an automatic capacity transition. The manual is Comcare operational guidance, not the Act, and licensees must map it to their own obligations, licence conditions and delegations. From Certificate Review Dates Are Workflow Triggers, Not Capacity Decisions

No. The guardrails ban liability language entirely. Liability for an injury is determined under section 14 of the SRC Act by the authorised decision maker using current legislation, policy, evidence and delegations. The skill organises the evidence record and flags gaps and conflicts. It never states, suggests or implies a view on liability or claim outcome. From A Claim Chronology Skill File Built for De-Identification

No. Comcare's aims to make contact within two business days and provide a formal written response within 15 business days describe Comcare's own service. A self-insured licensee needs its own approved complaint policy, ownership rules and response standards for the service lane. From A Complaint Is Not a Reconsideration: AI Can Route the Issue, Not Close It

Yes. Claim files hold names, claim numbers, dates of birth and detailed medical and personal information. None of that should go into an AI tool. Replace identifiers with placeholders, and only ask the model to work on the de-identified structure of the opinions. The comparison AI produces is a working aid, and the delegate applies it back to the real file. From Two Doctors Disagree: AI Can Map the Conflict, Not Resolve It

Yes. Reconsideration is the gateway. ART review of a primary determination requires a reconsideration first. Going straight from a primary determination to the ART is procedurally not available. The reconsideration produces a reviewable decision, which is the artefact that triggers the next review tier. From ART Review Rights Under the SRC Act: A Practitioner's Map

Yes, always and first. Replace the name with [CLAIMANT_NAME], the claim number with [CLAIM_NUMBER] and date of birth with [DATE_OF_BIRTH], mask other identifying detail, and keep the re-identification key separate. Use only an approved tool for sensitive data, never a public consumer model where input may be retained. From AI and Permanent Impairment: Organise the Evidence, Keep the Judgement

Yes. The public profile of a tool is not the same as a documented privacy assessment, so the toolkit applies regardless of vendor. Even where a tool claims no data leaves your environment, de-identify unless inference location, logging, training use, and Privacy Impact Assessment questions are all confirmed. The cost of de-identifying is low; being wrong is high. From The De-Identification Toolkit for Case Managers Working With AI

No. Regulation 11A(3) lists periods not counted for initial section 14 decisions while specified information or evidence is being obtained. Those pauses do not apply to a claimant's reconsideration request. New material belongs in the evidence lane of the matrix, but it does not create an AI-managed pause button for the reconsideration clock. From A Reconsideration Request Needs an Issue Map, Not an Outcome Prediction

Not automatically. Section 54(5) says strict compliance with an approved form is not required and substantial compliance is sufficient. Comcare's claim-form guidance deems licensee forms produced in accordance with that guidance to be approved, and advises authorities receiving another scheme's form to begin considering the information wherever feasible. A human must assess the form and substantial compliance. From Section 54 Intake: AI Can Find Gaps, Not Move the Clock

No. Where an examination is carried out, the written assessment under subsection 36(8) states the employee's capability of undertaking a rehabilitation program and, where appropriate, the kind of program. A section 37 determination that the employee should undertake a program is made separately by the rehabilitation authority, which must have regard to the section 36(8) assessment and the other matters listed in subsection 37(3). From AI Can Organise a Section 36 Rehabilitation Assessment. It Cannot Choose the Program

No. An offer of support does not prove section 14 liability. Accepting support is not an admission, and declining it is not contrary evidence. Treating participation, refusal, withdrawal or contact speed as a proxy for severity, credibility or causation converts an early-support program into an undisclosed liability screen. From Early Support Before Liability: Track the Process, Not the Person

No. The Comcare guidance is explicit that the calendar day count does not freeze when information is requested from the employee unless the request is made via a notice issued under section 58, and requests to an employer by other methods do not freeze the count either. A phone call or an email that is not a statutory notice leaves the clock running. From AI Can Run the Section 14 Clock. It Cannot Decide.

The OAIC says that where an entity creates personal information with reference to, or generated, inferred or observed from, other information it holds, this is a collection of personal information and APP 3 obligations apply. It gives artificial intelligence, automated decision making and data analytics as examples. A human privacy practitioner must decide which rules and exceptions apply to the particular record. From Early Support Before Liability: Track the Process, Not the Person

Broadly, yes, for the same injury. Under the framework in sections 46, 48 and 50, once a person recovers damages from a third party for the injury, the relevant authority can recover the compensation it has paid and further compensation is affected, with a limited exception for damages recovered for non-economic loss under a section 45 election. The precise effect depends on the facts, so the guidance and the Act should be read directly. From Preventing Double Payment Under the SRC Act: AI Can Flag the Overlap, Not Calculate the Offset

Clause 3.7 records that the documents a decision-maker is required to give the Tribunal on commencement of a review may include reports that fall within the scope of the Practice Direction. Where a party wants to place particular reliance on such a report but it does not carry the information in clauses 3.1 to 3.5C, the party should consider whether additional information should be sought from the expert and given to the Tribunal. The practical effect is that a report written well before any review is later read against the standard. From The Medical Report Now Has to Declare Its AI

No. The Tribunal is not concerned with whether AI tools were used in drafting. It is concerned with whether the determination is supported by evidence, whether the delegate can articulate the reasoning, and whether procedural fairness was observed. Where the reasoning trail is intact, AI use is unproblematic and effectively invisible at review. From Reading the Reasoning Trail: A Case Note on AI Drafted Determinations

As at 31 July 2026, no AI-specific wording was located in the current publicly published LCPM standard licence and LCPI templates or the CMS and RMS audit suite. That does not exclude a special condition, variation, direction or correspondence applying to a particular licensee. The SRCC's June 2026 AI Transparency Statement concerns the SRCC's own AI governance, not a licensee condition. From Your AI Register Is Not Licence Evidence Until It Maps to the SRCC Criteria

Set up a project space with a tight system prompt, a reference pack including the SRC Act 1988 and the review report, and a guardrail file. Then run four prompt patterns: triage the recommendations, compare them section by section, map operational impact, and scaffold a submission for legal to finish. From Leveraging AI to assist dissecting the SRC Act Review

AI can take de-identified medical reports and build a dated chronology tagged by author and body region, map each report to the approved Guide's structure, and flag gaps or inconsistencies. It can also draft a neutral file summary once a person has done the thinking. Every output is a checklist to verify, never a finding. From AI and Permanent Impairment: Organise the Evidence, Keep the Judgement

Claim material is sensitive information. Strip names, claim numbers, dates of birth and any identifier before a single line goes near a model, and work in placeholders such as [CLAIMANT_NAME], [CLAIM_NUMBER] and [DATE_OF_BIRTH]. If you cannot de-identify it, it does not go in. This is the precondition for everything else. From AI and the Reasonable Administrative Action Exclusion: Map the Actions, Keep the Judgement

Reports carry identifiers in letterheads, signature blocks, practice details, and file references. Apply the five-category de-identification toolkit in full, using placeholders like TREATING_PRACTITIONER, PRACTICE, CLAIMANT_NAME, CLAIM_NUMBER, INJURY_DATE, and CONDITION consistently. Never paste a report into any tool that has not been approved by your scheme operator. From Treating Practitioner Reports and AI: Where the Workflow Helps and Where It Hurts

Never enter real names, claim numbers, dates of birth, employee IDs, provider names, exact dates or identifying injury details into a public or unapproved AI tool. Use fictional scenarios and [PLACEHOLDER] fields. Removing a name is not enough if the remaining facts can still identify the person. If context cannot be written without identifying detail, do not proceed. From AI Can Draft Recovery Conversation Scripts, but the Listening Stays Human

De-identify before the model ever sees the matter, treating it as a fixed step, not a judgement under time pressure. Use placeholders like [CLAIMANT_NAME], [CLAIM_NUMBER] and [DATE_OF_BIRTH], and neutral tags such as Treating Practitioner A. Raw fields never leave the local environment; a human maps placeholders back at review time. From Build a WC Evidence Chronology Tool Without Outsourcing Judgement

Build the library from fictional scenarios, de-identified chronologies and placeholder fields rather than real claim material. Tell each prompt what not to do, ban legal conclusions, and require source separation. The second, more important control is that a human reviews every output before it is saved, sent or relied on. From Prompt Libraries Make WC AI Safer Only When Human Review Comes First

Review whenever new medical evidence arrives, at agreed checkpoints, and when duties, symptoms or concerns change. Ask whether the plan is still supported by current evidence, whether duties remain safe and genuinely available, whether consultation raised concerns, and whether the reviewer recorded changes made to AI-assisted drafts. From AI Can Organise Recovery-at-Work Information, but People Must Decide

A neat timeline is not evidence until a human checks it against the source records. Add an export gate that blocks export until a reviewer confirms in a required note that the source records have been checked. The export must carry a disclaimer that the chronology is a preparation artefact, not a determination. From Build a WC Evidence Chronology Tool Without Outsourcing Judgement

Section 19(2A)(b) measures cumulative hours the employee was prevented from working against 45 times normal weekly hours, and Comcare's FAQ says retrospective incapacity periods count. Comcare's section 116 guidance counts qualifying compensation leave in each seven-day block as a week, and those weeks need not be consecutive or align with the section 19 count. Copying one count into the other field can be precise and wrong. From Backdated Incapacity Needs Three Ledgers, Not One Payroll Fix

It removes the sorting so you can spend your time on the weighing. Working from a de-identified file, AI can lay out what clinical question each opinion addressed, the history and evidence each relied on, where the opinions agree, where they diverge, and where the record is missing something. That structured comparison makes the real points of difference obvious, which is where the delegate's judgement is best spent. From Two Doctors Disagree: AI Can Map the Conflict, Not Resolve It

Comcare's scheme guidance points to the Clinical Framework for the Delivery of Health Services, published by WorkSafe Victoria, as the lens for the cost-versus-benefit weighing, and notes the Administrative Review Tribunal has supported its use as a relevant consideration in determining reasonableness. Its five principles cover measured effectiveness, a biopsychosocial approach, empowering the injured person, goals focused on function and return to work, and the best available research evidence. From AI Can Build the Section 16 Picture, Not Make the Call

De-identification happens before input, using placeholders such as [CLAIMANT_NAME], [CLAIM_NUMBER] and [DATE_OF_BIRTH], with date tokens or offsets for the timeline. The file also carries a stop-and-ask guardrail. If the assistant detects apparent real identifiers in the input, it stops and asks for de-identified material instead of proceeding. From A Claim Chronology Skill File Built for De-Identification

Where AI assisted the original determination, the file note should record its role, the reconsideration reasons should engage with it, and the ART evidence pack should include the AI-assisted artefacts. The pattern that survives review is AI in the analysis, a human in the decision, documented in the reasoning trail. From ART Review Rights Under the SRC Act: A Practitioner's Map

Version the file. The maintenance prompt checks the statutory references against the compilation in force, currently No. 82, C2026C00285, in force 1 July 2026, checks the review rights wording, and recommends keep, update or retire. Any edit bumps the version line inside the file so drift is visible. From A Determination Evidence-Check Skill File Under Section 14

Add a short file note paragraph recording that AI was used, that inputs were de-identified, and that the case manager reviewed and edited the draft. Silence is not defensible. A clear note signals discipline and is read positively at review, while the absence of such a note is read negatively. From Reading the Reasoning Trail: A Case Note on AI Drafted Determinations

De-identification is the default for every workflow where claim data leaves the scheme's controlled environment. Remove full names, claim numbers, specific addresses, employer identifiers, and diagnoses linked to identifiers. Use stable internal identifiers the case manager can re-attach afterwards. This is the single highest-leverage control across the entire claims workflow. From AI Tools in Workers Compensation Claims: Where Value, Where Risk, Where Governance

No prompt sent to any external AI tool may contain a claimant name, claim number, date of birth, exact address, treating practitioner name, or employer reference. Replace them with placeholders such as CLAIMANT_NAME, CLAIM_NUMBER, CONDITION and TREATING_PRACTITIONER. This is treated as a control, not a guideline, every single time. From SRC Act and AI Assisted Determinations: A Practitioner Framework

Make the determination on the evidence and record the divergence in the file note, capturing the model's prediction, your view, and the basis for the difference. The override is logged in an override register, keeping the reasoning trail intact and protecting every individual determination while also generating data that improves the model. From Predictive Analytics and Claims Triage: A Risk Analysis for Scheme Operators

Use a consistent convention: square brackets, capital letters, no spaces, so AI outputs return in a form you can find and replace at the end. Common placeholders include [CLAIM_NUMBER], [CLAIMANT_NAME], [CONDITION], [INJURY_DATE], and [TREATING_PRACTITIONER]. Build any additional placeholders on the same pattern so each is self-explanatory. From The De-Identification Toolkit for Case Managers Working With AI

Name the four elements of the MI spirit, partnership, acceptance, compassion and evocation, and name OARS: open questions, affirmations, reflective listening and summaries. Impose constraints: no persuasion or pressure, no medical advice or recovery predictions, no commentary on liability or claim outcomes, plain Australian English. Unconstrained models default to upbeat salesmanship, the opposite of acceptance. From AI Can Draft Recovery Conversation Scripts, but the Listening Stays Human

AI should not infer attitude, motivation or credibility from an employee's language, or summarise concerns as resistance or non-compliance unless the evidence genuinely supports that and the decision-maker has considered context. Use AI to draft neutral prompts about whether duties are safe and sustainable, recognising section 37 considerations. From AI Can Organise Recovery-at-Work Information, but People Must Decide

Write a brief one-paragraph file note covering four things: the case manager calculated the figure manually, cross-checked it against an AI tool with de-identified inputs, the figures matched or any discrepancy was resolved, and the manual figure is authoritative. The note is not optional and is defensible at audit and review. From The Incapacity Cross-Check Workflow: AI as a Calculation Auditor

The case manager should redo the calculation independently, even where the AI got it right, so the figure is genuinely the case manager's. AI assisted maths accepted without the delegate being able to explain it creates calculation opacity. Use the AI result only as a cross-check, with the case manager's figure being authoritative. From Reading the Reasoning Trail: A Case Note on AI Drafted Determinations

Replace direct identifiers with merge fields such as [CLAIMANT_NAME] and assess indirect identifiers before anything reaches a model. The OAIC warns that removing names and addresses alone may not achieve de-identification because context and access affect re-identification risk. Treat a rare role, small work location or dated system incident as a possible indirect identifier, use an approved environment and preserve the original outside the model workflow. From Statement Summaries Need Source IDs, Not a Neat Story

No. Clause 1.11 says a failure to comply may have consequences for the weight the Tribunal gives to the expert's evidence. The note under it records that the Tribunal is not bound by the rules of evidence relating to opinion evidence, citing section 52 of the Administrative Review Tribunal Act 2024, and that the Tribunal will determine the weight given to any evidence before it. The report is not excluded. It can quietly carry less weight. From The Medical Report Now Has to Declare Its AI

Yes. Comcare's guidance on the amendments confirms that requests to undergo medical examinations under section 57 of the SRC Act are determinations that are reviewable and can be subject to review by the Administrative Review Tribunal, following changes that commenced on 14 June 2024. Because it is a determination for the purposes of section 60, notice of it must be given in writing under section 61 with the terms of the determination and the reasons for it, which raises the bar on how the referral is documented. From Section 57 Examinations: AI Can Build the Referral, Not Make the Call

No. A register can record the tool, purpose and owner while saying nothing about whether a workers compensation control operated on a given morning. Licence evidence links an approved use case to an executed licence clause, LCPM performance item or claims audit criterion, then to dated operating evidence, exceptions, a named human owner and an assurance result. From Your AI Register Is Not Licence Evidence Until It Maps to the SRCC Criteria

The evidence is promising but thin. A 2017 Canadian trial of 728 claimants found the MI group transitioned to modified duties more often with roughly half the benefit recurrence. A 2021 mapping review found a large research gap, and the 2022 MI-NAV trial cut sickness absence by about seven days, a difference that was not statistically significant. From AI Can Draft Recovery Conversation Scripts, but the Listening Stays Human

No. It is a search result, not proof of absence. Record the search scope with it. A result based on the claim-management system alone must not claim that payroll, correspondence and an approved provider channel were also checked. Treating a search miss as evidence of recovery is one of the prohibited transitions. From Certificate Review Dates Are Workflow Triggers, Not Capacity Decisions

No. It is a proposed internal control. The current LCPM embeds an annual Licensee Compliance and Performance Improvement report template covering prior-year performance, management-system activity, results, future strategies and executive certification, and permits a licensee's own branding and formatting. The crosswalk may support accurate LCPI inputs, but a human executive must be satisfied with any certification. From Your AI Register Is Not Licence Evidence Until It Maps to the SRCC Criteria

No. The case manager owns the calculation and runs the maths fully and personally, because section 19 is a delegated decision. AI is well suited to finding discrepancies between two computed figures, so it audits the manual figure rather than producing it. The manual figure stays authoritative every time. From The Incapacity Cross-Check Workflow: AI as a Calculation Auditor

No. The AI drafts the scaffold only: a submission outline, points keyed to recommendation numbers, evidence inputs needed, and points requiring legal advice. Government relations and external legal counsel finish the submission. A regulator can usually tell when a position came from the model rather than the organisation. From Leveraging AI to assist dissecting the SRC Act Review

The five categories are direct claimant identifiers, indirect identifiers, treating practitioner identifiers, third-party identifiers, and free-text leakage. They are deliberately broader than the strict legal definition of personal information, because the goal is robustness rather than minimum compliance. Replace each with a standard bracketed placeholder before pasting anything. From The De-Identification Toolkit for Case Managers Working With AI

Summary drift can subtly distort the practitioner's view, case managers may start reading the summary instead of the report, and privacy creep can leave identifiers in metadata. Audit failure modes include confirmation summarisation, compression of disagreement between treating and IME views, and loss of clinical specificity. Going back to the source is the key control. From Treating Practitioner Reports and AI: Where the Workflow Helps and Where It Hurts

Six categories arise in practice: procedural fairness drift, training data bias, reasoning trail dilution, vendor opacity, calibration decay, and concentration of effect where one model touches every claim. Each is real even where the model is good, and several only become visible in aggregate or once something goes wrong. From Predictive Analytics and Claims Triage: A Risk Analysis for Scheme Operators

Twenty calendar days for claims made in respect of an injury other than a disease, or an aggravation of an injury other than a disease. Sixty calendar days for claims made in respect of a disease. Thirty calendar days to decide a request by a claimant to reconsider a determination. The periods commenced on 1 April 2024 and apply to Comcare and self-insured licensees. From AI Can Run the Section 14 Clock. It Cannot Decide.

Under section 62, a request must set out reasons and reach the determining authority within 30 days after the determination first came to the requester's notice, unless a further period is allowed. For a claimant's request, regulation 11A(2) prescribes 30 calendar days to decide it, starting on the day the determining authority receives it, and receipt is day one even on a weekend or public holiday. From A Complaint Is Not a Reconsideration: AI Can Route the Issue, Not Close It

Under subsection 39(1), the employee must have an injury resulting in impairment and be undertaking or have completed a rehabilitation program, or have been assessed as not capable of undertaking one. The costs must be payable by the employee, and the alteration, modification, aid or appliance must be reasonably required, having regard to the impairment and, where appropriate, the rehabilitation-program requirements. From Section 39 Requests Need a Branching Evidence Map, Not One Checklist

They stop a person being compensated twice for the same injury. Comcare guidance SRC344 groups them into three overlaps: damages recovered from a third party at common law under sections 46, 48 and 50, an overlapping state or territory workers compensation entitlement under section 118, and compensation under a state or territory scheme that is not workers compensation under section 119. Each has its own recovery mechanism. From Preventing Double Payment Under the SRC Act: AI Can Flag the Overlap, Not Calculate the Offset

Five steps. Calculate the section 19 figure manually. De-identify the inputs with placeholders. Prompt the AI to compute the figure step by step. Compare the manual and AI figures. Document both on file, recording that the manual figure is authoritative. It adds five to ten minutes. From The Incapacity Cross-Check Workflow: AI as a Calculation Auditor

Observed document state, which holds only source-anchored facts. Unresolved evidence question, which states what the file does not establish and uses unknown rather than substituting zero, no restriction or full capacity. And human-owned action, which records an authorised next step, an owner and an internal due date. The third column may never say cease payment, close entitlement or return to pre-injury duties. From Certificate Review Dates Are Workflow Triggers, Not Capacity Decisions

Review runs in three tiers. Tier 1 is internal reconsideration under section 62, carried out by an uninvolved officer. Tier 2 is a merits review by the Administrative Review Tribunal under section 64. Tier 3 is a Federal Court appeal on a question of law under section 44 of the ART Act. From ART Review Rights Under the SRC Act: A Practitioner's Map

Reconsideration of a primary determination should generally be requested within 30 days of the determination being notified. An ART application must generally be lodged within 60 days of the reviewable decision being received. A Federal Court appeal must be lodged within 28 days. Extensions are available only in limited circumstances. From ART Review Rights Under the SRC Act: A Practitioner's Map

They have different triggers. Section 62(3) concerns when the request must be given, which is within 30 days after the day the determination first came to the requester's notice, unless the authority allows a further period. Section 62(6) and regulation 11A(2) require a claimant's request to be decided within 30 calendar days starting on the day the determining authority receives it. From A Reconsideration Request Needs an Issue Map, Not an Outcome Prediction

Gate one is senior management reviewing the high-impact recommendation triage before further analysis runs. Gate two is external or in-house legal counsel reviewing the section-by-section statutory comparison before the transitional and submission patterns run. Skipping either gate is the most common failure mode. From Leveraging AI to assist dissecting the SRC Act Review

Criterion, control, operating evidence, exception, owner, assurance result and reporting decision. Each row cites the exact source and version, states what must operate and what the AI is prohibited from doing, points to dated artefacts rather than a policy, records failures and overrides, names an accountable human role, records who tested the control, and captures the human notification assessment. From Your AI Register Is Not Licence Evidence Until It Maps to the SRCC Criteria

Build the three ledgers from immutable source extracts, join them by interval ID, and flag mismatches such as a claim period with no payroll match or a leave debit overlapping a determined period. It must not calculate incapacity amounts, decide leave entitlement, apply section 23A, propose a debt, restore leave or post any correction. Those decisions belong to claims, payroll and, where triggered, legal review. From Backdated Incapacity Needs Three Ledgers, Not One Payroll Fix

That two recorded values differ, or that a nominated search scope did not produce a match. Nothing more. A difference detected label and a no match located label both describe the records, not the world. Neither explains why the difference exists, and a search result is never proof of absence. The revised approval may simply sit outside the folder that was searched. From Provider Invoice Anomalies Are Queries, Not Allegations

It cannot catch inputs that are wrong on file; it will compute the wrong answer correctly. It cannot resolve legal characterisation, such as whether a payment is part of normal weekly earnings. It cannot make the choice of which available figures to use. The case manager remains responsible for these. From The Incapacity Cross-Check Workflow: AI as a Calculation Auditor

The tool organises facts: it orders events by date, tags evidence types, references sources, flags gaps and drafts review questions. The delegate decides outcomes under the SRC Act, including liability under section 14, injury under section 5A, disease under section 5B, incapacity, impairment, treatment and rehabilitation. From Build a WC Evidence Chronology Tool Without Outsourcing Judgement

The ART conducts an independent merits review and can affirm the reviewable decision, vary it, set it aside and substitute its own decision, or set it aside and remit the matter to the determining authority with directions. There are some limits on substitution, notably for certain rehabilitation decisions. From ART Review Rights Under the SRC Act: A Practitioner's Map

Three exact categories: an alteration to the employee's place of residence or place of work, a modification to a vehicle or article used by the employee, and an aid or appliance for the employee's use, with paragraph 39(1)(e) also providing for repair or replacement of an aid or appliance. The words are not to be widened into any purchase that could make daily life easier. From Section 39 Requests Need a Branching Evidence Map, Not One Checklist

Use only an approved tool and data pathway. De-identify before upload, using placeholders such as [CLAIMANT_NAME], [CLAIM_NUMBER], [CONDITION] and [DATE_OF_INJURY]. Removing a name alone is not enough if dates, providers, locations or unusual combinations of facts could still identify the person. From AI Can Organise a Section 36 Rehabilitation Assessment. It Cannot Choose the Program

Subsection 4(1) defines household services as services of a domestic nature, including cooking, house cleaning, laundry and gardening services, that are required for the proper running and maintenance of the employee's household. Whether a claimed task meets that definition depends on its nature and purpose, and the check belongs to the authorised decision-maker, not to a model. From AI Can Map a Section 29 Household Services Claim. It Cannot Decide What Is Reasonable

Do not enter real claimant names, claim numbers, employee IDs, provider names, exact dates, locations or medical details into unapproved AI tools. Use fictional examples and placeholder fields. Removing a name alone is not enough, because other facts can still identify the person. From AI Can Organise Recovery-at-Work Information, but People Must Decide

The statutory character of the request, not the item name. The subsection 4(1) definition of medical treatment includes the supply, replacement or repair of an artificial limb or other artificial substitute, or of a medical, surgical or other similar aid or appliance, feeding section 16. Comcare's guidance says items for treating the injury are not payable under section 39, which instead covers items assisting basic personal or work functions or supporting a return to work. Keyword rules are unsafe; the evidence decides. From Section 39 Requests Need a Branching Evidence Map, Not One Checklist

Section 8 governs normal weekly earnings, built from average weekly hours, the average hourly ordinary time rate of pay and qualifying allowances, with an overtime component under section 8(2) where the employee was required to work overtime on a regular basis. Section 9 identifies the relevant period, with the default under section 9(1) being the latest two-week period before the date of injury during which the employee was continuously employed by the Commonwealth or a licensed corporation. Section 19 then uses NWE within the incapacity compensation framework. From Normal Weekly Earnings: AI Can Assemble the Evidence, Not Set the Figure

That the practitioner proposed a review at that point. Comcare's medical evidence guidance says a certificate must contain a review date and not be open-ended, which makes the date a control for evidence currency. It is not a medical opinion about what happened after it. Nothing in the SRC Act gives a certificate date self-executing effect on capacity, entitlement, payments or work arrangements. From Certificate Review Dates Are Workflow Triggers, Not Capacity Decisions

Four things, walked through in order: the legal test that applied (which SRC Act section and threshold), the evidence considered, the reasoning connecting that evidence to the test, and the conclusion that issued. AI drafting can support every step but cannot replace any of them. The delegate must reproduce the trail on demand. From Reading the Reasoning Trail: A Case Note on AI Drafted Determinations

It audits a draft determination for completeness before the delegate decides. It extracts every factual assertion, maps each to a source document, checks the statutory elements under sections 5A, 5B and 14 are addressed with evidence, flags hedged wording and missing review rights text, and lists every gap for the delegate. From A Determination Evidence-Check Skill File Under Section 14

It sorts incoming claims by likely complexity, expected duration, or risk of dispute, supporting faster triage, load balancing, earlier intervention on escalating claims, and cleaner management reporting. A handful of Australian schemes run production deployments and many more are piloting, so the technology is now business as usual in pockets. From Predictive Analytics and Claims Triage: A Risk Analysis for Scheme Operators

Comcare's April 2026 guide says a statement may add information that did not fit in the claim form or respond to a request for further information, asks for factual and objective information, relevant dates, a helpful chronology and available supporting records, and tells employees the statement forms part of the claim file and will be released to the employer under section 59. From Statement Summaries Need Source IDs, Not a Neat Story

De-identification means more than removing a name. Claim numbers, employee IDs, provider names, exact dates, locations, unusual role details, injury facts, medical histories and combinations of facts can still identify a person. Use placeholder fields and fictional examples unless the tool, workflow and data handling have been formally approved. From Prompt Libraries Make WC AI Safer Only When Human Review Comes First

Comcare guidance is clear that being dissatisfied with the terms of a request does not amount to a reasonable excuse, and neither does a difficulty that arises only from the claimant's personal preference, such as choosing to relocate away from where the requirement can be met. Practical difficulties that are objectively reasonable, such as a remote location, an accessibility need or not receiving the request, are more likely to qualify. From Reasonable Excuse Under the SRC Act: AI Can Structure the Analysis, Not Make the Finding

Three things. The treatment must have been obtained in relation to the compensable injury, it must have been reasonable for the employee to obtain in the circumstances, and compensation is of such amount as the authority determines is appropriate to that treatment. Before any of that, the request must fall within the definition of medical treatment in section 4(1). Something outside that definition is not section 16 medical treatment at all. From AI Can Build the Section 16 Picture, Not Make the Call

Section 5A excludes compensation for an injury suffered as a result of reasonable administrative action taken in a reasonable manner in respect of the employee's employment. The analysis runs in order: first whether there is a disease under section 5B, then whether any employment causative factor was reasonable administrative action across four questions, then whether the injury was suffered as a result of that action. From AI and the Reasonable Administrative Action Exclusion: Map the Actions, Keep the Judgement

Clause 3.5B requires an expert preparing a written report for a Tribunal proceeding to state whether the report includes content generated by using Generative AI. Clause 3.5C requires, where it does, that the expert clearly identify the AI content and the applications used to generate it, and certify that they have personally checked all of it, including research and other material cited in support, and are satisfied it is all accurate and reliable. Clause 3.5A defines Generative AI and names ChatGPT, Gemini, Microsoft Copilot, Perplexity, Claude, Grok and DeepSeek AI as examples. From The Medical Report Now Has to Declare Its AI

Comcare's material on the Guide states that relevant authorities must consider an employee's circumstances, rely as much as possible on relevant information from the injured employee's treating practitioner, and seek and take into account the employee's views about the selection of the person or people who will conduct the assessment or examination. Those are judgement calls about a specific person, which is exactly why they stay with the delegate and are documented in the decision record. From Section 57 Examinations: AI Can Build the Referral, Not Make the Call

Nine fields: a stable assertion ID, a source anchor (document, page and paragraph), the speaker as a placeholder, a neutral close paraphrase, the knowledge type (direct observation, reported account, inference, opinion or unclear), the date status, any linked source the speaker named, a descriptive corroboration status, and a precise question for human review. One row per discrete assertion, never a merged claim. From Statement Summaries Need Source IDs, Not a Neat Story

The exact operative terms, including entitlement, amount, period and effective date; each material finding and the evidence relied on; the statutory provision, test and legal terms used; the treatment of conflicting evidence; authorised role names; and the applicable reconsideration, Comcare-review or ART pathway with its time statement and contact details. Each item carries an anchor. From Plain-English Determinations Need a Reader Test, Not New Reasons

Six controls: a written model description before deployment, ongoing calibration monitoring reviewed quarterly at minimum, an override register reviewed for patterns, bias monitoring across cohorts, reasoning trail integration in file notes, and a Privacy Impact Assessment covering training data, production flow, and output, reviewed annually. From Predictive Analytics and Claims Triage: A Risk Analysis for Scheme Operators

Comcare's guidance says the SRC Act does not expressly deal with leave accrual during the pre-determination period. Accrual of National Employment Standards leave in that period is governed by the Fair Work Act 2009 and instruments made under it, and section 130 of that Act generally restricts taking or accruing NES leave during a workers compensation period unless a compensation law permits it. From Backdated Incapacity Needs Three Ledgers, Not One Payroll Fix

An authorised human must make and communicate it. Section 60 of the SRC Act includes a section 16 decision within the definition of a determination, and section 61 generally requires written notice of the terms, reasons and reconsideration route. Subsection 61(2) disapplies that section for a determination under subsection 16(1) that compensation equal to the full amount of the cost is payable to a person other than the employee. Section 62 provides the reconsideration pathway. From Provider Invoice Anomalies Are Queries, Not Allegations

The Regulations do not prescribe a deemed determination or a reviewable decision where a determining authority fails to meet the statutory timeframe. Comcare guidance states that determining authorities should have regard to other legislation that may enable a claimant to seek review of a determination not made within the prescribed timeframes. Because the timeframes are not determinations under section 60(1), they are not subject to reconsideration under section 62 or appeal to the Administrative Review Tribunal, and an application would be made to the Federal Court of Australia. From AI Can Run the Section 14 Clock. It Cannot Decide.

Comcare's statutory-timeframes guidance states the Regulations do not create a deemed determination or reviewable decision merely because the period expires. The claim still requires urgent human action, escalation and a lawful determination, and the intake record should show exactly where the time went. From Section 54 Intake: AI Can Find Gaps, Not Move the Clock

It is a short, reusable markdown instruction file that captures exactly how a claims team builds a chronology. It covers purpose, triggers, required inputs, method, output format and guardrails. Installed as ChatGPT project instructions, a Claude skill or a Copilot agent, it applies the team standard on every run instead of relying on each person to re-prompt it. From A Claim Chronology Skill File Built for De-Identification

It is the ground a claimant relies on when they fail to comply with a requirement, such as attending an examination or undertaking a rehabilitation program. Comcare guidance SRC345 sets a combined test: the delegate weighs both the subjective grounds the claimant advances and whether those grounds are objectively reasonable. The excuse must show the person was unable to comply, not merely unwilling, and it requires more than a rational explanation. From Reasonable Excuse Under the SRC Act: AI Can Structure the Analysis, Not Make the Finding

A proposed control that sits above the complaint summary and survives every hand-off: the original message ID and unaltered receipt timestamp, the exact decision or event identified, the candidate lanes with their trigger words, the statutory clock source, a separate human owner for each lane and confirmation that no lane is treated as resolved by another. From A Complaint Is Not a Reconsideration: AI Can Route the Issue, Not Close It

Use four stages. Select an approved prompt from the library. Insert only de-identified placeholder facts or fictional scenario facts. Review the output against source documents, legislation, policy and tone requirements. Then record the AI assistance, reviewer, sources checked and changes made before the content is saved or sent. From Prompt Libraries Make WC AI Safer Only When Human Review Comes First

Under section 57 of the SRC Act, a delegate can require an employee to undergo an examination by a medical practitioner where additional medical information or specialist opinion is needed to make a decision. The resulting independent opinion often sits alongside the treating doctor's report, and the two can disagree. A decision to require a section 57 examination is itself a reviewable determination for the purposes of section 60. From Two Doctors Disagree: AI Can Map the Conflict, Not Resolve It

A proposition-by-proposition comparison of the approved capsule and the edited notice. It returns the before text, after text, protected anchor, difference type and human disposition, treating moved text as a presentation change while still flagging altered wording. A changed effective date, omitted qualifier or rewritten evidence finding fails, and an authorised human resolves it. From Plain-English Determinations Need a Reader Test, Not New Reasons

A separately labelled set of files added to examine a known control concern, decision type or site. Targeted files can reveal a control failure, but they must not be blended into the representative core, and a clean targeted file cannot repair weak core coverage. Keep the selection basis immutable for every file. From A Defensible Claims Audit Sample Starts With the Population, Not an AI Risk Score

A control boundary for AI in NWE work. The AI may extract, reconcile, sort and flag records, but it may not total earnings, apply the section 8 formula, nominate the relevant period, classify overtime as required and regular, or decide that an allowance qualifies. Wherever legal characterisation is needed, the output says human classification required, and every factual entry carries a file and page, row or cell reference. From Normal Weekly Earnings: AI Can Assemble the Evidence, Not Set the Figure

An issue-to-ground matrix uses one row for each challenged term. Each row records the challenged term, the original reason in the determination, the evidence relied on, the ground supplied by the requester, any alleged omission, new material with its stated proposition, the requested outcome and exact source coordinates. Separate rows stop a general paragraph from concealing which reason relates to which requested change. From A Reconsideration Request Needs an Issue Map, Not an Outcome Prediction

A control that displays, beside every process status, the meanings that status does not carry. DECLINED means the employee did not accept the offered support; it does not mean uncooperative, recovered or low severity. WITHDRAWN means participation ended; it does not explain why. Making the prohibited readings visible is more reliable than expecting users to remember them. From Early Support Before Liability: Track the Process, Not the Person

Open the source document, save a clearly marked working copy, run a five-category Find and Replace sweep for names, dates, locations, practitioner details, and third parties, then visually scan paragraph by paragraph for anything missed. Once clean, use the working copy in the approved tool while the original stays untouched in the source system. From The De-Identification Toolkit for Case Managers Working With AI

Administrative action is directed specifically at the employee and their employment relationship, such as a performance appraisal, counselling, suspension or disciplinary action under section 5A(2). Operational action, a direction about how and when to perform the work itself, is not reasonable administrative action, and any resulting injury is compensable. The distinction decides whether section 5A is even in play. From AI and the Reasonable Administrative Action Exclusion: Map the Actions, Keep the Judgement

Section 118 deals with an overlap between SRC Act compensation and state or territory workers compensation for the same injury, and Comcare may recover the amount it paid. Section 119 deals with an overlap between SRC Act compensation and a state or territory scheme that pays compensation other than workers compensation, such as a motor accident scheme, where Comcare is generally liable only for the excess. From Preventing Double Payment Under the SRC Act: AI Can Flag the Overlap, Not Calculate the Offset

Frame the question and identify the legal test. De-identify the input with placeholders. Generate a structured draft. Map the draft to the actual claim file line by line. Then issue the decision, applying the legal test and signing as the delegated decision maker. AI assists only at step three. From SRC Act and AI Assisted Determinations: A Practitioner Framework

Triage with the full report first, then generate a structured summary as a navigation aid. Cross-check the summary against the original and correct any drift. Use the summary as an aid, not a replacement, when deciding under section 16 or section 14, and capture medical opinion in the practitioner's own words from the source. From Treating Practitioner Reports and AI: Where the Workflow Helps and Where It Hurts

Section 57A of the SRC Act requires Comcare to prepare a Guide for Arranging Rehabilitation Assessments and Requiring Examinations. The object of the Guide is to support ethical, transparent and accountable decision making in relation to arranging a rehabilitation assessment of an employee under subsection 36(1), or requiring an employee to undergo an examination under subsection 36(3) or 57(1), including appropriate consideration of the employee's personal circumstances. It is a legislative instrument, compliance is mandatory, and it applies to section 36 and section 57 determinations made on or after 30 October 2024. From Section 57 Examinations: AI Can Build the Referral, Not Make the Call

Section 24 generally requires the permanent impairment to reach at least 10 per cent whole person impairment, with specific exceptions for matters such as hearing loss and loss of the use of fingers, toes, or the sense of taste or smell. The 10 per cent line is legal; whether the evidence reaches it is a clinical judgement. From AI and Permanent Impairment: Organise the Evidence, Keep the Judgement

The Type of Occurrence Classification System is the primary Australian system for classifying work-related injury and disease incidents across workers compensation systems and for national work health and safety statistics. Safe Work Australia identifies the current version as Edition 3 Revision 2, TOOCS3.2, with its catalogue data last updated in July 2025. From TOOCS Coding Is Data Quality, Not a Liability Finding

Three classes most often. Transposition errors, such as numbers in the wrong order or misplaced decimal points. Statutory interpretation errors, often involving section 8 normal weekly earnings interacting with allowances or prescribed amounts. Sequencing errors, where the order adjustments are applied matters for the legal correctness of the determination. From The Incapacity Cross-Check Workflow: AI as a Calculation Auditor

An unsafe prompt invites real claim information, asks AI to form a conclusion, and risks claimant-facing language without legal review. A safer prompt is narrower: it uses fictional information and placeholder fields, bans liability recommendations, separates evidence, gaps, assumptions and actions, and marks the output as draft-only for human review. From Prompt Libraries Make WC AI Safer Only When Human Review Comes First

Where relevant: the likely period of need; access and movement at home or work; vehicle access, driving, movement and safety; alternative transport; hire arrangements; and, where a place of residence or vehicle that received a compensated alteration or modification is later disposed of, any resulting increase in its value. The human decision-maker decides which matters carry weight and what amount is reasonable. From Section 39 Requests Need a Branching Evidence Map, Not One Checklist

Comcare guidance sets out notification duties. A claimant or dependant who decides to pursue a common law claim for damages against a third party must tell the relevant authority in writing as soon as practicable and within 7 days of becoming aware of the claim, and once damages are recovered must notify the authority in writing of the amount within 28 days. Recovering damages for the same injury affects the right to further SRC Act compensation. From Preventing Double Payment Under the SRC Act: AI Can Flag the Overlap, Not Calculate the Offset

Subsection 29(2) requires the decision-maker to have regard to five matters: the employee's pre-injury and post-injury contribution to household services, the people living in the household with their ages and needs, those members' contribution before the injury, what household or family members might reasonably be expected to provide for themselves and the employee after the injury, and the need to avoid substantial disruption to their employment or other activities. The subsection is expressed without limiting other matters, so other relevant circumstances may also be considered. From AI Can Map a Section 29 Household Services Claim. It Cannot Decide What Is Reasonable

Everything identifying: the claimant's name, claim number, date of birth and any detail that could identify them, replaced with placeholders such as [CLAIMANT_NAME], [CLAIM_NUMBER] and [CONDITION]. De-identification is mandatory before any claim material reaches a model, and the material goes only into a tool your organisation has approved for claims work. If no tool is approved, that approval comes first. From AI Can Build the Section 16 Picture, Not Make the Call

It must not invent or select grounds, assess credibility, apply a legal test, rank issues, estimate prospects or recommend affirming, varying or revoking the determination. It also must not draft an ART application or predict what the Tribunal might do. Where the request is unclear, it writes HUMAN CLARIFICATION REQUIRED rather than filling the gap. From A Reconsideration Request Needs an Issue Map, Not an Outcome Prediction

The reviewer verifies every extracted statement against its source, corrects dates and context, confirms de-identification, applies the current Act and Guide, preserves the employee-view and assessor-selection steps, and ensures only qualified and authorised people perform the assessment and determination functions. From AI Can Organise a Section 36 Rehabilitation Assessment. It Cannot Choose the Program

Four recipient tasks: what exactly was decided, why that result was reached, what happens next, and how the decision can be challenged. Each answer returns FOUND, AMBIGUOUS, NOT FOUND or CONFLICT, and every FOUND result requires an exact paragraph or page anchor. An answer without an anchor is a failure, however plausible it sounds. From Plain-English Determinations Need a Reader Test, Not New Reasons

The whole person impairment assessment against the Guide, weighing conflicting examiner evidence, combining multiple impairments by the combined values method, section 27 non-economic loss assessment, and the section 24 determination all stay human. A model can flag divergence, but conclusions, percentages and the entitlement decision must be attributable to qualified people. From AI and Permanent Impairment: Organise the Evidence, Keep the Judgement

For an audit of the entire claims management system and its implementation, the October 2025 Claims management system audit tool, version 5.0, suggests all files for populations of 1 to 15 claims, 16 files for 16 to 90, 30 for 91 to 150, 56 for 151 to 280, 70 for 281 to 500, and 100 files for 501 or more, with possible increases where multiple sites are tested. From A Defensible Claims Audit Sample Starts With the Population, Not an AI Risk Score

Enough for the report to answer clause 3.1 on its face: the questions or issues the examiner is asked to address, and a reference to the documents and materials provided. Comcare's guidance on engaging a legally qualified medical practitioner already says proper instructions, including the appropriate scope of the examination, reduce the risk of the examiner straying into areas irrelevant to the claim. Adding a request that the examiner address Generative AI use expressly costs a sentence and closes the gap before it opens. From The Medical Report Now Has to Declare Its AI

Never send unredacted claim data to an undocumented tool, treat an AI summary as a substitute for the source documents in a load-bearing decision, let an AI-drafted determination letter leave without case manager review, act on AI-flagged patterns without verifying them against sample claims, or run AI-influenced decisions with no file note record. From AI Tools in Workers Compensation Claims: Where Value, Where Risk, Where Governance

Five gates: a version gate recording the applicable TOOCS version and table release, an existence gate confirming each code and description pair exists in that version, a proof gate requiring a source phrase and anchor for each field, a relationship gate testing cross-field consistency, and a change gate preserving the full correction history. Every anomaly returns as a human-review question, never a replacement. From TOOCS Coding Is Data Quality, Not a Liability Finding

AI must not weigh competing medical evidence, apply the section 16 reasonableness test, perform causation analysis under section 14 and section 5B, or judge credibility on disputed history. These require human judgement with the rigour the SRC Act expects. Any AI output that opines on reasonableness or causation is out of scope. From Treating Practitioner Reports and AI: Where the Workflow Helps and Where It Hurts

A person goes and fills them. If the file holds no functional measure, no stated goal or no current treating practitioner review, the answer is a request back to the practitioner, not a model estimate. The work-up is valuable precisely because it names the gaps instead of papering over them. Asking the model to fill a gap converts missing evidence into invented evidence. From AI Can Build the Section 16 Picture, Not Make the Call

Gather combined evidence first: the current medical certificate, treating practitioner guidance, supervisor input on real and available duties, and employee feedback, plus review history. Only then use AI to draft a view of possible duties, restrictions, hazards, review questions and evidence gaps for human checking. From AI Can Organise Recovery-at-Work Information, but People Must Decide

Do not paste sensitive internal or Cabinet-in-confidence material into a shared endpoint without confirming data classification rules. Do not let AI draft the submission itself. Do not assume the model has read the report; enforce the citation rule. And do not run the workflow once and stop, since consultation is iterative. From Leveraging AI to assist dissecting the SRC Act Review

Capture the AI's role at the time of decision, not retrospectively. Record which tool was used, what data was sent to it, what the output was, and how the case manager engaged with it. Retrospective reconstruction of the AI's role is significantly harder than capture at the moment of decision. From AI Tools in Workers Compensation Claims: Where Value, Where Risk, Where Governance

A four-line standard holds up at audit: the AI tool and version used, the specific purpose the AI served, confirmation that inputs were de-identified, and a statement that the case manager reviewed and edited the output so the issued text reflects their own reasoning. Four lines, defensible at audit. From SRC Act and AI Assisted Determinations: A Practitioner Framework

Section 61(1A) requires the determining authority to consider and determine each section 14 claim within the period prescribed by the regulations. Under regulation 11A that period is 20 calendar days for an injury other than a disease and 60 calendar days for a disease, and day one is the day the determining authority receives a claim made in accordance with section 54. Comcare's guidance confirms the receipt day counts even on a weekend or public holiday. From Section 54 Intake: AI Can Find Gaps, Not Move the Clock

Five situations, and only for section 14 initial claims. A requirement under section 57 that the claimant undergo an examination by a nominated legally qualified medical practitioner. A request to the claimant under section 58 for information or a document. A requirement on an employer's principal officer under section 71. The claimant advising they will provide further evidence. The determining authority considering it reasonable and necessary to obtain further medical evidence by requiring a report from the claimant's practitioner or a nominated one. None of these apply to reconsideration requests. From AI Can Run the Section 14 Clock. It Cannot Decide.

Section 11A of the Regulations provides that the count commences on the date the determining authority receives either an initial claim made in accordance with section 54 of the SRC Act, or a request for reconsideration made by a claimant under section 38 or 62. Section 54 requires a written claim on a Comcare approved form and a certificate by a legally qualified medical practitioner, so a claim arriving without the certificate does not start the count until the certificate is received. The day of receipt is day one even if it is a Saturday, Sunday or public holiday. From AI Can Run the Section 14 Clock. It Cannot Decide.

Section 8(5) operates where, because of the shortness of the relevant period, the calculated result would not fairly represent the weekly rate at which the employee was being paid. It permits calculation over another period considered reasonable for reaching a fair representation. It is not a general licence to choose the most favourable, convenient or statistically stable period, and the delegate decides whether the statutory trigger is met. From Normal Weekly Earnings: AI Can Assemble the Evidence, Not Set the Figure

After a reviewable decision, not as a substitute for section 62 reconsideration of a primary determination. Section 63 requires the reviewable decision notice to state that application may be made to the ART, section 64 identifies who may apply, and section 65(4) requires an application within 60 days after the person is served with notice, with extensions available under section 19 of the Administrative Review Tribunal Act 2024. From A Complaint Is Not a Reconsideration: AI Can Route the Issue, Not Close It

Section 54 requires a written claim and, except for claims under sections 16 or 17, a certificate from a legally qualified medical practitioner. If the certificate is required but absent, section 54(3) treats the claim as not made until it is supplied. The receipt history should still preserve the original submission so the later certificate can be linked to it. From Section 54 Intake: AI Can Find Gaps, Not Move the Clock

The plain-English pass starts only after the human decision and reasons are complete and the decision-maker is identified. If AI spots a missing finding, an unsupported step or unexplained conflicting evidence, the document returns to the decision-maker. The model does not patch the gap with a more persuasive sentence, and it never completes reasoning during copy-editing. From Plain-English Determinations Need a Reader Test, Not New Reasons

Three conditions justify pausing the model: calibration drift exceeding tolerance so predictions are no longer reliably informative, a systematic bias finding that is not promptly addressable, and vendor changes to the underlying behaviour that the operator has not yet assessed. A paused model can be switched back on once issues are resolved. From Predictive Analytics and Claims Triage: A Risk Analysis for Scheme Operators

Avoid AI drafting for significant degree disease claims under section 5B, credibility-driven determinations, and determinations involving the interaction of multiple Act provisions such as sections 14, 16, 19 and 24 together. In these, review cost outweighs drafting savings. AI is still useful for narrower sub-tasks like summarisation. From SRC Act and AI Assisted Determinations: A Practitioner Framework

The Guide for Arranging Rehabilitation Assessments and Requiring Examinations 2024 requires the rehabilitation authority to first consider whether existing information is sufficient. If it is insufficient or inconsistent, further information or clarification should be sought from the employee's treating practitioner, and the authority should rely on the treating practitioner as much as possible before involving an independent assessor. The authorised person applies the Guide to the circumstances and records the decision. From AI Can Organise a Section 36 Rehabilitation Assessment. It Cannot Choose the Program

Re-identify only at the final write step, never as an intermediate one. All review and editing happens with placeholders intact, which keeps any discussion privacy safe. When ready to write the final text, run a controlled find and replace from placeholders back to real values inside your case management system, not in the AI tool. From The De-Identification Toolkit for Case Managers Working With AI

AI operates across five production workflows in Australian schemes: intake and initial triage, decision support on liability, communications drafting, document analysis and case file review, and quality assurance with pattern detection. Each carries genuine value alongside specific risks, so each needs its own governance baseline rather than a single blanket control. From AI Tools in Workers Compensation Claims: Where Value, Where Risk, Where Governance

The tool defines the eligible population as claims with some form of activity within the 12 months before the audit date. The first AI task is reconciling the frozen extract against activity records, sites and administrators, showing inclusions, exclusions, duplicates and missing fields. A competent human auditor then approves the population and scope. From A Defensible Claims Audit Sample Starts With the Population, Not an AI Risk Score

Section 60 defines a determination as a determination, decision or requirement made under a listed set of SRC Act provisions. A notice of intention to make a determination, or an administrative decision such as a deemed date of injury or the label of an accepted condition, does not become reconsiderable because a model labels it so. A human checks that threshold first. From A Reconsideration Request Needs an Issue Map, Not an Outcome Prediction

Suspicious provider, fraud risk, invalid service and excessive treatment. Replace them with neutral labels tied to observable records: possible duplicate identifier, approval match not located, service date outside recorded range, rate version review and provider field difference. The label should describe the field difference and nothing else. From Provider Invoice Anomalies Are Queries, Not Allegations

The same six-part file is platform neutral. It installs natively where a plan supports a Skills feature, runs as project instructions plus knowledge files in ChatGPT and Claude projects on other plans, and adapts to Microsoft 365 Copilot by pasting it into an agent's Instructions field in Agent Builder. From A Claim Chronology Skill File Built for De-Identification

It depends on the governing provision, decision-maker and stage, so it is never selected from a generic template. Section 61 generally requires terms, reasons and a statement about requesting reconsideration under subsection 62(2). Subsection 38(1) applies to a section 36 or 37 determination made by a rehabilitation authority other than a relevant authority. Section 63 requires an ART-rights statement. From Plain-English Determinations Need a Reader Test, Not New Reasons

SRC345 applies the same concept across sections 36(4), 37(7), 57(2), 58 and 118. Suspension for failing a section 36 rehabilitation examination, a section 37 program or a section 57 medical examination is made under sections 36(4), 37(7) or 57(2) and is a determination. A reasonable-excuse assessment on a refusal to deal with a claim under section 58, or a suspension under section 118, is not a determination. From Reasonable Excuse Under the SRC Act: AI Can Structure the Analysis, Not Make the Finding

Three provisions generate most difficult calls. Section 36 covers assessing an employee's capability of undertaking a rehabilitation program. Section 37 covers providing rehabilitation programs and considers the employee's attitude. Section 40 places a duty on the employer to take all reasonable steps to provide suitable employment. Conversations about what counts as suitable are where trust is most easily lost. From AI Can Draft Recovery Conversation Scripts, but the Listening Stays Human

Section 108B permits a licence to authorise the licensee, or a specified person acting on its behalf, to manage claims. Section 108C requires an authorised licensee to determine the claims it manages within the scope of its licence. Section 108D allows the Safety, Rehabilitation and Compensation Commission to grant a licence on conditions it considers necessary. Section 108E includes functions to make payments and determine claims accurately and quickly. From Your AI Register Is Not Licence Evidence Until It Maps to the SRCC Criteria

The SRCC Commission Data Warehouse Specifications version 3.0 still lists TOOCS3.1 for claims with a received by claims processing date on or after 1 July 2008, while also permitting the latest version for any claim. Preserve the receipt date, recorded version and current manual, then confirm the live warehouse implementation with Comcare before changing or resubmitting data. From TOOCS Coding Is Data Quality, Not a Liability Finding

Block terms that imply a determination the delegate has not made: accepted, rejected, liable, unreasonable and non-compliant. They suggest decisions reserved under sections 14 and 5A. Use neutral alternatives instead, such as event recorded pending review, gap or question flagged for delegate, and management action flagged for s5A review. From Build a WC Evidence Chronology Tool Without Outsourcing Judgement

Section 39 names the relevant authority, defined for an employee of a licensee as the licensee. For an employee of a licensed corporation the rehabilitation authority is the principal officer, and Comcare's guidance says that authority can assess whether the employee is not capable of undertaking a program. Subsections 4(10) and 4(10A) substitute the licensed authority or corporation for provisions outside the listed exclusions. Neither a case manager nor AI can invent that decision authority. From Section 39 Requests Need a Branching Evidence Map, Not One Checklist

A human decision-maker. Comcare's scheme guidance says the decision-maker is responsible for determining the figure, and that an initial NWE determination should identify the period used, explain why it was selected and summarise the data considered. A section 8 decision is a determination under the Act, so the notification and reconsideration provisions require a human process that states the terms and reasons and preserves review rights. From Normal Weekly Earnings: AI Can Assemble the Evidence, Not Set the Figure

Comcare monitors licence compliance on behalf of the Safety, Rehabilitation and Compensation Commission. Reviews may occur during the developing phase, at years two or six of a licence, or in response to a concern. The SRCC lists the audit tool, workbook and report template for audits after a self-insurance licence has been granted. From A Defensible Claims Audit Sample Starts With the Population, Not an AI Risk Score

Because it collapses three different questions: what compensation was determined and paid, what payroll processed, and what leave was debited or accrued under the governing law or instrument. The SRC Act deals with incapacity compensation and leave through different provisions, so a single reversal instruction can be precise in payroll terms and still wrong in legal terms. From Backdated Incapacity Needs Three Ledgers, Not One Payroll Fix

The four hierarchical structures are Nature, Bodily Location, Mechanism and Agency, applied in that order. The Agency structure supplies two linked outputs: the Breakdown Agency, the object, substance or circumstance most closely associated with the point at which things started to go wrong, and the Agency of Injury, what was directly involved in inflicting the injury or disease. Collapsing the two destroys useful information. From TOOCS Coding Is Data Quality, Not a Liability Finding

AI can format notes but must not become the source of truth. A reviewer must verify that every factual statement is traceable to a source document, that SRC Act references and review rights are accurate, and that no legal or medical conclusion has been invented. These are requirements AI cannot be left to approximate. From Prompt Libraries Make WC AI Safer Only When Human Review Comes First

Because the review pathway differs. A suspension under sections 36(4), 37(7) or 57(2) is a determination that can be reconsidered and then reviewed by the Administrative Review Tribunal, so the reasons must withstand that scrutiny. A decision under sections 58 or 118 is not a determination and is not reconsidered, but it can be challenged by judicial review, so procedural fairness and a clear written record still matter. From Reasonable Excuse Under the SRC Act: AI Can Structure the Analysis, Not Make the Finding

Determination outcomes use deterministic language. A determination states what is decided and the evidence basis for it. The skill lints the draft for hedging terms in outcome statements, including should, likely, probably and it appears, and flags each one so the delegate can replace it with a definite statement. From A Determination Evidence-Check Skill File Under Section 14

The review produced 124 recommendations, not legislation. Submissions, consultation forums and stakeholder input happen on the recommendation set, not the bill. An employer who waits for the bill is six to twelve months too late, so forming a coherent organisational position during consultation is the priority. From Leveraging AI to assist dissecting the SRC Act Review

Compression erases the boundaries between direct observation, reported account, inference and opinion, and it creates a verification tax: a reviewer who reads "the allocation changed after the meeting" must rediscover who asserted it, which meeting was meant and whether another account differed. A summary that reads well can quietly resolve conflicts the delegate was supposed to decide. From Statement Summaries Need Source IDs, Not a Neat Story

Because the rate column is a reference point, not a verdict. Comcare's scheme guidance says the appropriate amount is determined case by case by the relevant authority, describes the guidance as non-binding, and recognises that a higher fee could be paid where, for example, the isolated location at which the treatment is provided means the provider incurs additional costs. A difference from a published rate opens an evidence question rather than answering one. From Provider Invoice Anomalies Are Queries, Not Allegations

AI tools tend to write with assertive, confident prose that reads well at the desk but can read as overstatement at review. Over-confident wording that asserts more than the evidence supports is a long-standing review risk. Every confident assertion must be earned by evidence on file, or the paragraph should be rewritten. From Reading the Reasoning Trail: A Case Note on AI Drafted Determinations

Because allied health rates are reviewed three times a year, with updates taking effect from 1 February, 1 July and 1 November. Comparing every invoice against today's displayed rate manufactures false exceptions for services delivered earlier. Store the rate source, jurisdiction, profession, item, effective date and retrieval date, then compare against the service date rather than the invoice date. From Provider Invoice Anomalies Are Queries, Not Allegations

The reasonableness test is not prescriptive. There may be more than one reasonable way to take an action, and the question is whether what was done was reasonable in the circumstances. That is a contextual, evidence-weighing judgement that depends on what the employer knew at the time and on the particular employee. A determination has to survive reconsideration and Administrative Review Tribunal review on the strength of the human reasoning. From AI and the Reasonable Administrative Action Exclusion: Map the Actions, Keep the Judgement

A prompt library is a box of tools the user picks from, pastes and adapts each time, and it still earns its place for one-off tasks. A skill file is a standing instruction set the assistant applies every time the task appears, so the de-identification rules and the output standard cannot be forgotten on a busy day. From A Claim Chronology Skill File Built for De-Identification

Because the form records them separately and collapsing them destroys information. The Certificate of Capacity form records the examination date, the date of issue, the certified capacity category with its own applicable dates, an estimated return-to-work date and a review date. A single certificate expiry field cannot tell you whether the practitioner examined the person recently, what period the stated capacity actually covers, or when your own workflow deadline falls. From Certificate Review Dates Are Workflow Triggers, Not Capacity Decisions