--- title: "Insurance Law" url: "https://ailawframework.com/practice/insurance-law" type: "practice_area" site: "AI Law Framework" id: "A22" task_count: 24 last_reviewed: "2026-10-10" --- # Insurance Law > Read the actual policy and endorsements alongside the claim facts. Check coverage, exclusions, notifications and time limits; keep preliminary extraction separate from the lawyer’s coverage analysis. Canonical page: https://ailawframework.com/practice/insurance-law Last reviewed: 2026-10-10 Site index for agents: https://ailawframework.com/llms.txt Also covers: General insurance, life insurance, reinsurance, claims law, coverage disputes, policy interpretation, insurance litigation, professional indemnity insurance, D&O insurance ## Key facts - Tasks: 24 - Green: 4, Yellow: 17, Red: 3 - Levels: Green is AI drafts and the lawyer checks; Yellow is lawyer and AI draft together; Red is the lawyer writes and AI may only check, clarify or collect data. ## Principles - **Define the matter**: Confirm the client’s instructions, jurisdiction, intended output and the information available before dividing the work. - **Verify the authorities**: Open each relied-on authority independently. Confirm its existence, current status, proposition and any quotation or pinpoint reference. - **Reconcile the figures**: Trace every relied-on figure to its source, confirm dates and assumptions, and independently recalculate totals or derived amounts. ## Checklist - Open each relied-on authority independently. Confirm its existence, current status, proposition and any quotation or pinpoint reference. - Trace every relied-on figure to its source, confirm dates and assumptions, and independently recalculate totals or derived amounts. - Calculate operative dates independently from the current rule and actual triggering event. Check exceptions and record the verified deadline. - Check the current court or regulator requirements, prescribed form and attachments for the intended document and filing. - Review operative clauses in the complete document, including definitions, cross-references, schedules and inconsistencies with the instructions. - Confirm that the actual tool, feature and processing terms permit use of this information, including access, retention, training and disclosure conditions. - Obtain and inspect the relevant current official register or search result. Confirm identities, interests and the date and scope of the search. - Verify the applicable jurisdiction and the version of each relevant law, rule or instrument in force at the material time. - Resolve exceptions and record the responsible lawyer’s approval of the exact output and intended use. ## Restrictions - Confirm firm, client and applicable court or professional restrictions before using any AI feature. ## Jurisdiction note Apply this guidance to the actual matter and jurisdiction. Confirm current requirements before relying on an example. ## Basis Practice guidance supports planning; the responsible lawyer confirms the scope, evidence and review requirements for the matter. Also covers: General insurance, life insurance, reinsurance, claims law, coverage disputes, policy interpretation, insurance litigation, professional indemnity insurance, D&O insurance Insurance law in Australia spans policy drafting and interpretation under the Insurance Contracts Act 1984 (Cth) (ICA), claims handling, coverage disputes, subrogation, professional indemnity, reinsurance, and regulatory compliance under the ASIC Act, Life Insurance Act 1995, and APRA's prudential standards. The AI risk profile is high: coverage questions turn on precise policy language and fact-specific exclusion analysis where hallucinated authorities or mistaken statutory references cause immediate client harm; claims and litigation work involves tight deadlines, jurisdiction-specific procedural rules, and AFCA/court filing obligations where errors are not correctable without cost or prejudice; and the sector's own rapid adoption of AI in underwriting, claims triage, and fraud detection creates a concurrent layer of governance advice risk under CPS 230, ICA s.13, and the incoming Privacy Act APP 1.7 automated decision-making disclosure obligations (effective 10 December 2026). ## Limits on AI assistance Red lines — never in this area ## Limits on AI assistance AI must never be the sole or unverified source of case citations submitted to a court, AFCA, or included in a client advice letter — every authority must be confirmed on AustLII, Westlaw AU, or LexisNexis AU before any use. Handa & Mallick [2024] FedCFamC2F 957 (Victorian disciplinary action against 'Mr Dayal') and Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (NSW OLSC referral) are the established Australian enforcement precedents. AI must never generate final insurance policy wording, endorsements, or exclusions that are transmitted to a client, counterparty, or insured without the lawyer reconstructing from a verified precedent and reviewing the complete document — AI-generated policy wording goes directly into binding contracts. AI must never draft or reconstruct affidavits, witness statements, character references, or expert reports — prohibited in NSW (SC Gen 23, effective 3 February 2025, extended to District Court and Land and Environment Court), required to be disclosed and verified in Victoria (SC Gen 25, effective 14 May 2026), and subject to candour obligations in the Federal Court (GPN-AI, issued 16 April 2026); one AI tool cannot be used to verify the output of another (SC Gen 25 requirement). AI must never determine the applicable limitation period for an insurance claim or subrogation action without independent legal verification — the ICA, State Limitation Acts, and policy-specific limitation clauses interact and AI frequently mis-states the operative period. AI must never be used to generate a coverage position (coverage or no coverage) that is sent to a client or counterparty without full lawyer review of the actual policy wording — AI paraphrases of exclusion clauses and defined terms are the highest-risk failure mode in coverage advice. AI must never assess whether an insurer's conduct breaches the duty of utmost good faith (ICA s.13) without a lawyer applying the specific facts to the current statutory and Code regime — ASIC v TAL Life Limited [2021] FCA 193 establishes that the duty requires individualised assessment of each claim file. AI must never generate a non-disclosure or misrepresentation analysis for a consumer insurance contract without the lawyer confirming whether the post-5 October 2021 regime (s.20B duty to take reasonable care not to make a misrepresentation, replacing the former s.21A) applies — applying the superseded duty to a consumer contract is likely to produce legally incorrect advice. Confidential client information, claim details, medical records, or policy schedules must not be entered into a public or uncontracted LLM — insurance claim files routinely contain sensitive health, financial, and business information engaging Privacy Act (Cth) APP 3 and APP 11 obligations and ASCR r.9 confidentiality duties; Hamilton Locke (2025) identifies privilege waiver through public AI tool use as a live risk. AI must never produce a regulatory compliance position regarding CPS 230, ASIC RG 271, or Privacy Act APP 1.7 automated decision-making for an insurer or insured without the lawyer independently confirming the current version of each instrument — CPS 230 was amended 30 April 2026 with effect 1 July 2026 and APP 1.7 commences 10 December 2026. AI must never be relied upon as the basis for an insurer client's decision to deny or reduce an insurance claim — an algorithmic denial without human review of the individual claim file is inconsistent with ICA s.13 utmost good faith as established in ASIC v TAL Life Limited and creates a civil penalty risk under ICA s.13(2A). AI must never produce reinsurance treaty analysis as if Australian ICA principles apply without the lawyer first confirming the governing law clause — the majority of international reinsurance treaties are governed by English or New York law and the ICA does not apply. - AI must never be the sole or unverified source of case citations submitted to a court, AFCA, or included in a client advice letter — every authority must be confirmed on AustLII, Westlaw AU, or LexisNexis AU before any use. Handa & Mallick [2024] FedCFamC2F 957 (Victorian disciplinary action against 'Mr Dayal') and Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (NSW OLSC referral) are the established Australian enforcement precedents. - AI must never generate final insurance policy wording, endorsements, or exclusions that are transmitted to a client, counterparty, or insured without the lawyer reconstructing from a verified precedent and reviewing the complete document — AI-generated policy wording goes directly into binding contracts. - AI must never draft or reconstruct affidavits, witness statements, character references, or expert reports — prohibited in NSW (SC Gen 23, effective 3 February 2025, extended to District Court and Land and Environment Court), required to be disclosed and verified in Victoria (SC Gen 25, effective 14 May 2026), and subject to candour obligations in the Federal Court (GPN-AI, issued 16 April 2026); one AI tool cannot be used to verify the output of another (SC Gen 25 requirement). - AI must never determine the applicable limitation period for an insurance claim or subrogation action without independent legal verification — the ICA, State Limitation Acts, and policy-specific limitation clauses interact and AI frequently mis-states the operative period. - AI must never be used to generate a coverage position (coverage or no coverage) that is sent to a client or counterparty without full lawyer review of the actual policy wording — AI paraphrases of exclusion clauses and defined terms are the highest-risk failure mode in coverage advice. - AI must never assess whether an insurer's conduct breaches the duty of utmost good faith (ICA s.13) without a lawyer applying the specific facts to the current statutory and Code regime — ASIC v TAL Life Limited [2021] FCA 193 establishes that the duty requires individualised assessment of each claim file. - AI must never generate a non-disclosure or misrepresentation analysis for a consumer insurance contract without the lawyer confirming whether the post-5 October 2021 regime (s.20B duty to take reasonable care not to make a misrepresentation, replacing the former s.21A) applies — applying the superseded duty to a consumer contract is likely to produce legally incorrect advice. - Confidential client information, claim details, medical records, or policy schedules must not be entered into a public or uncontracted LLM — insurance claim files routinely contain sensitive health, financial, and business information engaging Privacy Act (Cth) APP 3 and APP 11 obligations and ASCR r.9 confidentiality duties; Hamilton Locke (2025) identifies privilege waiver through public AI tool use as a live risk. - AI must never produce a regulatory compliance position regarding CPS 230, ASIC RG 271, or Privacy Act APP 1.7 automated decision-making for an insurer or insured without the lawyer independently confirming the current version of each instrument — CPS 230 was amended 30 April 2026 with effect 1 July 2026 and APP 1.7 commences 10 December 2026. - AI must never be relied upon as the basis for an insurer client's decision to deny or reduce an insurance claim — an algorithmic denial without human review of the individual claim file is inconsistent with ICA s.13 utmost good faith as established in ASIC v TAL Life Limited and creates a civil penalty risk under ICA s.13(2A). - AI must never produce reinsurance treaty analysis as if Australian ICA principles apply without the lawyer first confirming the governing law clause — the majority of international reinsurance treaties are governed by English or New York law and the ICA does not apply. ## Jurisdiction-specific considerations Australian traps ## Jurisdiction-specific considerations ICA 1984 is Commonwealth legislation but State courts apply it; insurance litigation may be in State Supreme Courts, the Federal Court, or AFCA — each has different AI practice notes and disclosure requirements that must be confirmed for each forum. Federal Court GPN-AI practice note (issued 16 April 2026): if AI was used in preparing pleadings, submissions, evidence, or chronologies, the responsible lawyer must personally verify facts are provable, cited authorities exist and support the propositions advanced, and evidence is admissible. Where disclosure is required or ordered, it must appear at the start of the document. All persons must be able to inform the Court what AI was used, how, and for what purpose. NSW Supreme Court Practice Note SC Gen 23 (effective 3 February 2025, extended to NSW District Court and Land and Environment Court): AI-generated content prohibited in affidavits, witness statements, and character references without court leave. Expert reports in NSW courts must comply with the Expert Witness Code of Conduct — the AI use restrictions in SC Gen 23 extend to expert report preparation instructions. Victoria Supreme Court Practice Note SC Gen 25 (issued 14 May 2026, replacing the 2024 Guidelines for Litigants): a formal practice note (not merely guidelines) applicable to all Victorian Supreme Court users. AI use is permitted but practitioners must identify specific AI-generated portions, explain how output was verified, and note that one AI tool cannot be used to verify another AI tool's output. Victorian practitioners must treat this as binding. Queensland Practice Direction 5 of 2025 (Supreme Court) and QCAT Practice Direction 10 of 2025: both address accuracy of references in submissions and impose obligations on practitioners to verify any AI-generated references before filing. Handa & Mallick [2024] FedCFamC2F 957 ('Mr Dayal' case): Victorian VLSB+C disciplined the practitioner (prohibited from practising as a principal, handling trust money, or practising unsupervised for two years) for filing unverified AI-generated case citations in a family law matter. Ignorance of a tool's limitations was not accepted as a defence — technological literacy is now part of the duty of competence. Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95: NSW lawyer admitted filing submissions containing 17 non-existent cases generated by ChatGPT; referred to the NSW OLSC. Demonstrates that the Handa & Mallick enforcement posture extends to all Australian jurisdictions. AFCA monetary limits are adjusted triennially, not annually: the most recent adjustment took effect 1 January 2024 (general insurance complaint threshold $1,263,000; compensation cap $6,300 for general insurance per claim). The next scheduled adjustment is 1 January 2027. AI training data will reflect a prior set of figures — always verify on the AFCA Rules current at time of complaint lodgment. ICA consumer insurance reform (effective 5 October 2021): the former s.21A (life insurance disclosure duty) was repealed and replaced by s.20B (duty to take reasonable care not to make a misrepresentation) for Consumer Insurance Contracts. AI trained before or near that date regularly applies the superseded s.21A duty to consumer contracts. The reform shifted the burden to the insurer to ask the questions it needs. CPS 230 (APRA Operational Risk Management Standard): effective 1 July 2025; targeted amendments released 30 April 2026, revised standard and CPG 230 effective 1 July 2026. AI vendors used in insurance claims, underwriting, or fraud detection are potential 'material service providers' requiring formal risk management — the definition must be checked against the current Attachment to CPS 230. Non-traditional service providers (including some AI platform providers) may qualify for a limited contractual exemption under the April 2026 amendments. Privacy Act 1988 (Cth) APP 1.7 automated decision-making disclosure obligations: effective 10 December 2026 — insurers using AI for claims triage, underwriting scoring, fraud flagging, or premium calculation affecting individuals must disclose this in their privacy policy. OAIC guidance is expected by September 2026. Legal advice on AI-driven insurance operations must address this incoming obligation and the compliance gap for current systems. General Insurance Code of Practice: currently under comprehensive revision by the Insurance Council of Australia; public consultation occurred Q1 2026, Code expected to be lodged with ASIC for approval mid-2026. The current 2021 Code remains in force but specific provisions may be superseded. Check the Insurance Council of Australia website before relying on any specific Code provision in advice. ASIC v Auto & General Insurance Company Limited — first UCT decision in insurance: primary decision [2024] FCA 272 (March 2024) found no unfair contract term; Full Federal Court [2025] FCAFC 76 (June 2025) upheld the insurer's position but on a different reasoning, implying a materiality threshold into the notification clause. The Full Court reasoning is the current operative precedent on UCT in insurance — AI will not know the appeal outcome. Motor accident and workers compensation schemes are State-based and vary significantly (SIRA in NSW, TAC in Victoria, WorkCover Queensland, ICWA in Western Australia, ReturnToWorkSA) — AI has poor reliability in distinguishing State scheme rules and frequently conflates them with the Commonwealth ICA regime. Never rely on AI for jurisdiction-specific scheme entitlements or time limits. ASIC s.13 ICA (utmost good faith) civil penalty: ICA s.13(2A) (inserted March 2019) makes breach of the duty of utmost good faith by an insurer a civil penalty provision enforceable by ASIC. AI-generated claims handling recommendations that result in unfair denials or delays expose the client insurer to penalty proceedings as illustrated by ASIC v TAL Life Limited [2021] FCA 193 and ASIC v OnePath Life (proceedings commenced 2022). - ICA 1984 is Commonwealth legislation but State courts apply it; insurance litigation may be in State Supreme Courts, the Federal Court, or AFCA — each has different AI practice notes and disclosure requirements that must be confirmed for each forum. - Federal Court GPN-AI practice note (issued 16 April 2026): if AI was used in preparing pleadings, submissions, evidence, or chronologies, the responsible lawyer must personally verify facts are provable, cited authorities exist and support the propositions advanced, and evidence is admissible. Where disclosure is required or ordered, it must appear at the start of the document. All persons must be able to inform the Court what AI was used, how, and for what purpose. - NSW Supreme Court Practice Note SC Gen 23 (effective 3 February 2025, extended to NSW District Court and Land and Environment Court): AI-generated content prohibited in affidavits, witness statements, and character references without court leave. Expert reports in NSW courts must comply with the Expert Witness Code of Conduct — the AI use restrictions in SC Gen 23 extend to expert report preparation instructions. - Victoria Supreme Court Practice Note SC Gen 25 (issued 14 May 2026, replacing the 2024 Guidelines for Litigants): a formal practice note (not merely guidelines) applicable to all Victorian Supreme Court users. AI use is permitted but practitioners must identify specific AI-generated portions, explain how output was verified, and note that one AI tool cannot be used to verify another AI tool's output. Victorian practitioners must treat this as binding. - Queensland Practice Direction 5 of 2025 (Supreme Court) and QCAT Practice Direction 10 of 2025: both address accuracy of references in submissions and impose obligations on practitioners to verify any AI-generated references before filing. - Handa & Mallick [2024] FedCFamC2F 957 ('Mr Dayal' case): Victorian VLSB+C disciplined the practitioner (prohibited from practising as a principal, handling trust money, or practising unsupervised for two years) for filing unverified AI-generated case citations in a family law matter. Ignorance of a tool's limitations was not accepted as a defence — technological literacy is now part of the duty of competence. - Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95: NSW lawyer admitted filing submissions containing 17 non-existent cases generated by ChatGPT; referred to the NSW OLSC. Demonstrates that the Handa & Mallick enforcement posture extends to all Australian jurisdictions. - AFCA monetary limits are adjusted triennially, not annually: the most recent adjustment took effect 1 January 2024 (general insurance complaint threshold $1,263,000; compensation cap $6,300 for general insurance per claim). The next scheduled adjustment is 1 January 2027. AI training data will reflect a prior set of figures — always verify on the AFCA Rules current at time of complaint lodgment. - ICA consumer insurance reform (effective 5 October 2021): the former s.21A (life insurance disclosure duty) was repealed and replaced by s.20B (duty to take reasonable care not to make a misrepresentation) for Consumer Insurance Contracts. AI trained before or near that date regularly applies the superseded s.21A duty to consumer contracts. The reform shifted the burden to the insurer to ask the questions it needs. - CPS 230 (APRA Operational Risk Management Standard): effective 1 July 2025; targeted amendments released 30 April 2026, revised standard and CPG 230 effective 1 July 2026. AI vendors used in insurance claims, underwriting, or fraud detection are potential 'material service providers' requiring formal risk management — the definition must be checked against the current Attachment to CPS 230. Non-traditional service providers (including some AI platform providers) may qualify for a limited contractual exemption under the April 2026 amendments. - Privacy Act 1988 (Cth) APP 1.7 automated decision-making disclosure obligations: effective 10 December 2026 — insurers using AI for claims triage, underwriting scoring, fraud flagging, or premium calculation affecting individuals must disclose this in their privacy policy. OAIC guidance is expected by September 2026. Legal advice on AI-driven insurance operations must address this incoming obligation and the compliance gap for current systems. - General Insurance Code of Practice: currently under comprehensive revision by the Insurance Council of Australia; public consultation occurred Q1 2026, Code expected to be lodged with ASIC for approval mid-2026. The current 2021 Code remains in force but specific provisions may be superseded. Check the Insurance Council of Australia website before relying on any specific Code provision in advice. - ASIC v Auto & General Insurance Company Limited — first UCT decision in insurance: primary decision [2024] FCA 272 (March 2024) found no unfair contract term; Full Federal Court [2025] FCAFC 76 (June 2025) upheld the insurer's position but on a different reasoning, implying a materiality threshold into the notification clause. The Full Court reasoning is the current operative precedent on UCT in insurance — AI will not know the appeal outcome. - Motor accident and workers compensation schemes are State-based and vary significantly (SIRA in NSW, TAC in Victoria, WorkCover Queensland, ICWA in Western Australia, ReturnToWorkSA) — AI has poor reliability in distinguishing State scheme rules and frequently conflates them with the Commonwealth ICA regime. Never rely on AI for jurisdiction-specific scheme entitlements or time limits. - ASIC s.13 ICA (utmost good faith) civil penalty: ICA s.13(2A) (inserted March 2019) makes breach of the duty of utmost good faith by an insurer a civil penalty provision enforceable by ASIC. AI-generated claims handling recommendations that result in unfair denials or delays expose the client insurer to penalty proceedings as illustrated by ASIC v TAL Life Limited [2021] FCA 193 and ASIC v OnePath Life (proceedings commenced 2022). ## Risks to address How it goes wrong ## Risks to address Hallucinated ICA case authority: AI generates plausible-sounding Australian insurance case names (e.g. 'AAI Limited v XYZ Pty Ltd [2023] NSWSC 123') that do not exist — submitted to AFCA or a court, this is a professional conduct breach as established in Handa & Mallick [2024] FedCFamC2F 957 (Victorian VLSB+C disciplinary action: practitioner banned from principal practice for two years) and Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (NSW OLSC referral for 17 non-existent ChatGPT-generated cases). ICA s.54 / s.57 conflation: AI routinely confuses the 'act or omission' discretion in s.54 (which the High Court has interpreted broadly) with the prejudice-to-insurer defence, producing coverage advice that is wrong on the law and misleading to the client. Consumer insurance disclosure duty error: AI applies the pre-5 October 2021 duty of disclosure to consumer insurance contracts (general or life) rather than the post-2021 duty under s.20B (reasonable care not to make a misrepresentation), producing incorrect non-disclosure advice that may cause a practitioner to advise an insurer to avoid a contract when avoidance is legally unsustainable. State scheme conflation: AI merges Victorian TAC, NSW SIRA, Queensland WorkCover, and ICWA rules into a single description of workers compensation or motor accident entitlements — highly dangerous in jurisdiction-specific claim advice, particularly for limitation periods and benefits entitlements. Stale AFCA jurisdictional thresholds: AI provides AFCA monetary limits from a superseded adjustment period — the threshold of $1,263,000 effective 1 January 2024 will not change until 1 January 2027 but AI may hold figures from the prior triennial period; complaint jurisdictional advice based on wrong figures may incorrectly direct a client to or away from AFCA. Stale regulatory data on CPS 230 and GI Code: AI provides requirements from a superseded version of CPS 230 (pre-July 2025 effective date or pre-July 2026 amended standard) or from the current GI Code rather than the forthcoming revised Code; regulatory compliance advice to an insurer client may be wrong at the threshold level. - Hallucinated ICA case authority: AI generates plausible-sounding Australian insurance case names (e.g. 'AAI Limited v XYZ Pty Ltd [2023] NSWSC 123') that do not exist — submitted to AFCA or a court, this is a professional conduct breach as established in Handa & Mallick [2024] FedCFamC2F 957 (Victorian VLSB+C disciplinary action: practitioner banned from principal practice for two years) and Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (NSW OLSC referral for 17 non-existent ChatGPT-generated cases). - ICA s.54 / s.57 conflation: AI routinely confuses the 'act or omission' discretion in s.54 (which the High Court has interpreted broadly) with the prejudice-to-insurer defence, producing coverage advice that is wrong on the law and misleading to the client. - Consumer insurance disclosure duty error: AI applies the pre-5 October 2021 duty of disclosure to consumer insurance contracts (general or life) rather than the post-2021 duty under s.20B (reasonable care not to make a misrepresentation), producing incorrect non-disclosure advice that may cause a practitioner to advise an insurer to avoid a contract when avoidance is legally unsustainable. - State scheme conflation: AI merges Victorian TAC, NSW SIRA, Queensland WorkCover, and ICWA rules into a single description of workers compensation or motor accident entitlements — highly dangerous in jurisdiction-specific claim advice, particularly for limitation periods and benefits entitlements. - Stale AFCA jurisdictional thresholds: AI provides AFCA monetary limits from a superseded adjustment period — the threshold of $1,263,000 effective 1 January 2024 will not change until 1 January 2027 but AI may hold figures from the prior triennial period; complaint jurisdictional advice based on wrong figures may incorrectly direct a client to or away from AFCA. - Stale regulatory data on CPS 230 and GI Code: AI provides requirements from a superseded version of CPS 230 (pre-July 2025 effective date or pre-July 2026 amended standard) or from the current GI Code rather than the forthcoming revised Code; regulatory compliance advice to an insurer client may be wrong at the threshold level. ## Review checklist The lawyer's review checklist ## Review checklist Every case citation in AI output verified as real, correctly named, and currently good law on AustLII (cases.austlii.edu.au) or Westlaw AU — including check for subsequent history (appeal, overrule, disapproval) — before any internal or external use. AFCA determinations must be verified separately on the AFCA published decisions database; they are not on AustLII. Every statutory reference checked against the current version on the Federal Register of Legislation (legislation.gov.au) — ICA, Life Insurance Act, Privacy Act, Corporations Act, and ASIC Act have all been materially amended; specifically confirm whether s.20B (consumer insurance misrepresentation) or the pre-2021 disclosure regime applies to the matter at hand. Policy wording reviewed in full from the actual policy document (including all endorsements and schedules), not AI-generated paraphrase — AI summaries of exclusion clauses and defined terms strip the ambiguity that is the coverage question. Time limits verified independently: claim limitation period in the specific policy; applicable State Limitation Act; AFCA complaint timeframes; RG 271 IDR response obligations; court filing deadlines — do not rely on AI for any date-critical obligation. Client confidentiality and privilege check: confirm no sensitive client data, claim file contents, health records, or financial information was entered into an uncontracted or public AI tool; if data was input, assess Privacy Act APP breach notification obligations and whether privilege may have been waived (Hamilton Locke 2025 analysis). Jurisdiction and applicable law check: confirm whether the matter is governed by the ICA (general/life insurance), Marine Insurance Act 1909, a State motor accident or workers compensation scheme, or an international reinsurance treaty under English or New York law — AI defaults to the ICA regardless of the applicable law. Court AI disclosure compliance: if the matter is in the Federal Court (GPN-AI, April 2026), NSW Supreme Court, District Court, or Land and Environment Court (SC Gen 23, February 2025), or Victorian Supreme Court (SC Gen 25, May 2026), confirm whether any AI-generated content requires disclosure at the head of the document, and that the verification standard has been met — Victoria SC Gen 25 expressly prohibits using one AI tool to verify another. AI-generated draft documents reviewed for internal consistency across all sections — AI produces inconsistent defined terms, contradictory exclusion logic, and mis-matched recitals when drafting policies, deeds, or multi-section submissions in multiple passes. Pay particular attention to cross-references between defined terms and operative exclusion clauses. - Every case citation in AI output verified as real, correctly named, and currently good law on AustLII (cases.austlii.edu.au) or Westlaw AU — including check for subsequent history (appeal, overrule, disapproval) — before any internal or external use. AFCA determinations must be verified separately on the AFCA published decisions database; they are not on AustLII. - Every statutory reference checked against the current version on the Federal Register of Legislation (legislation.gov.au) — ICA, Life Insurance Act, Privacy Act, Corporations Act, and ASIC Act have all been materially amended; specifically confirm whether s.20B (consumer insurance misrepresentation) or the pre-2021 disclosure regime applies to the matter at hand. - Policy wording reviewed in full from the actual policy document (including all endorsements and schedules), not AI-generated paraphrase — AI summaries of exclusion clauses and defined terms strip the ambiguity that is the coverage question. - Time limits verified independently: claim limitation period in the specific policy; applicable State Limitation Act; AFCA complaint timeframes; RG 271 IDR response obligations; court filing deadlines — do not rely on AI for any date-critical obligation. - Client confidentiality and privilege check: confirm no sensitive client data, claim file contents, health records, or financial information was entered into an uncontracted or public AI tool; if data was input, assess Privacy Act APP breach notification obligations and whether privilege may have been waived (Hamilton Locke 2025 analysis). - Jurisdiction and applicable law check: confirm whether the matter is governed by the ICA (general/life insurance), Marine Insurance Act 1909, a State motor accident or workers compensation scheme, or an international reinsurance treaty under English or New York law — AI defaults to the ICA regardless of the applicable law. - Court AI disclosure compliance: if the matter is in the Federal Court (GPN-AI, April 2026), NSW Supreme Court, District Court, or Land and Environment Court (SC Gen 23, February 2025), or Victorian Supreme Court (SC Gen 25, May 2026), confirm whether any AI-generated content requires disclosure at the head of the document, and that the verification standard has been met — Victoria SC Gen 25 expressly prohibits using one AI tool to verify another. - AI-generated draft documents reviewed for internal consistency across all sections — AI produces inconsistent defined terms, contradictory exclusion logic, and mis-matched recitals when drafting policies, deeds, or multi-section submissions in multiple passes. Pay particular attention to cross-references between defined terms and operative exclusion clauses. ## Tasks (24) | Task | Level | How AI is used | What the lawyer must review | | --- | --- | --- | --- | | Review the policy wording and draft a coverage advice table | Yellow | Extract insuring clauses, exclusions, conditions, notification and limits into a table mapped to the notified facts, with clause references and open questions. | Construe the wording against proven facts yourself Do not send a generated table as a coverage position | | Draft claims-handling advice from the policy and the claim file | Yellow | From the claim correspondence and the policy, draft an advice outline (notification, prejudice, reservation of rights, next steps) with document cites. | Decide whether to decline, reserve or indemnify (with the insurer client) Check claims-handling standards in the applicable jurisdiction | | [Draft an ombudsman or external-dispute submission](https://ailawframework.com/practice/insurance-law/draft-an-ombudsman-or-external-dispute-submission.md) | Yellow | Structure an ombudsman or external-dispute submission from the claim chronology, with a list of published determinations cited for you to check. | Confirm monetary jurisdiction and time limits Remember the forum applies fairness, not only strict law Verify every cited determination in the forum's own… | | Draft policy endorsements and exclusion wording from the underwriting brief | Yellow | On the lawyer's instructions, assist with the endorsement or exclusion clause, working from the underwriter's brief, the base wording it amends and the… | The lawyer settles the operative words, because an ambiguity in an exclusion is read against the insurer.… | | Map subrogation and contribution rights from the policy and the loss file | Yellow | From the policy, settlement and third-party documents, table possible recovery targets, waivers of subrogation and other-insurance clauses. | Decide whether a right survives settlement Check the release before the insurer pays | | Draft pleadings in a coverage or indemnity dispute | Red | The lawyer drafts the statement of claim, defence, reply, cross-claim or particulars, and the tool writes no part of it — no draft, no first draft, no… | The lawyer drafts the pleading and settles every allegation, admission and non-admission personally: an allegation not traversed is deemed admitted on filing,… | | Draft expert instructions for a loss assessor, engineer or actuary | Red | Take the questions and the assumed facts the lawyer has settled and set them out in a working table for the lawyer's own use, with the policy clause and the… | The lawyer writes the letter of instruction and settles every assumed fact in it. Check each question against the coverage issue actually in dispute, confirm… | | [Map the insurer's conduct, licensing and claims-code obligations](https://ailawframework.com/practice/insurance-law/map-the-insurer-s-conduct-licensing-and-claims-code-obligations.md) | Yellow | From the client's claims process and policies, table candidate conduct, licensing and claims-code obligations with a list of current instruments to check. | Apply the current regulator and code text Do not advise from a stale summary; automated-decision rules are moving | | [Build a non-disclosure or misrepresentation table from application and claim](https://ailawframework.com/practice/insurance-law/build-a-non-disclosure-or-misrepresentation-table-from-application-and-c.md) | Yellow | Compare the application answers with later facts on the claim file and table possible non-disclosure or misrepresentation issues with document cites. | Confirm whether a duty of disclosure or a duty not to misrepresent applies Apply the different remedies for innocent and fraudulent cases Do not treat the… | | [List AI-use risk points in the client's claims or underwriting process](https://ailawframework.com/practice/insurance-law/list-ai-use-risk-points-in-the-client-s-claims-or-underwriting-process.md) | Yellow | From the client's process description, list discrimination, explainability, disclosure and claims-handling risk points and questions for the vendor. | Treat this as client advice, not a software assessment of the firm's tools Apply the current conduct rules yourself | | Review a reinsurance treaty and extract the claims-cooperation clauses | Yellow | Extract attachment, exclusions, claims-cooperation, follow-the-settlements/fortunes and aggregation clauses into a table with clause references. | Get specialist input on placement and wording disputes Do not advise a cedent or reinsurer from a generated summary of a treaty slip | | Draft a settlement deed and release for an insurance claim | Yellow | From the agreed heads, draft a deed covering payment, release, confidentiality, subrogation and any coverage reservation. | Check who is released; a broad release can kill subrogation Check whether other policies are affected | | Draft a coverage-position letter from the approved issues table | Yellow | Turn the lawyer-approved coverage table into a first-draft position letter (indemnify / reserve / decline) with policy cites and information requests. | Decide the position with the client Check timing and prejudice arguments, which are easy to get wrong Do not send a generated decline | | Build a notification and prejudice chronology from the claim correspondence | Yellow | From the policy's notification clause and the correspondence, table when the insured knew, when notice was given, and any alleged prejudice, with document… | Decide late-notification consequences; they are legal and fact-specific Do not treat the chronology as a declinature | | Compare the claim narrative with the contemporaneous documents | Yellow | Compare the insured's claim form and statements with incident records, contracts and correspondence; list inconsistencies and missing attachments. | Remember inconsistencies may be innocent Leave fraud allegations to the lawyer and the client | | Extract other-insurance, allocation and deductible mechanics across a programme | Green | From the tower or programme wordings, table limits, deductibles, drop-down, other-insurance and drop-in clauses with conflicts flagged. | Get specialist input on allocation across years and towers Confirm complete wordings, including endorsements | | Summarise a claims file and policy for a coverage review | Yellow | Extract facts, dates and policy terms into a coverage memo outline. | Lawyer forms the coverage view. | | Analyse claims-handling records for AI-driven decisions | Yellow | Identify where automated tools affected the claim decision. | Lawyer assesses duty of utmost good faith and GICOP. | | Build a chronology of a contested insurance claim | Green | Create a dated chronology from emails and reports. | Verify against documents. | | Review policy wording for AI-related exclusions | Yellow | Flag silent AI and cyber exclusions. | Lawyer interprets. | | Draft IDR response to a complainant | Red | Draft a response citing policy terms. | Lawyer settles; AFCA rules apply. | | Review insurer claims-handling AI against the AI joint statement | Yellow | Gap-analyse. | Lawyer advises. | | Summarise expert reports in an insurance claim | Green | Summarise. | Verify. | | Draft a policy for a firm's AI use on insurer panel work | Green | Draft. | Approve. | ## Jurisdiction notes ### Australia #### Scope Policy drafting and interpretation under the Insurance Contracts Act 1984 (Cth) (ICA), claims handling, coverage disputes, subrogation, professional indemnity and reinsurance. #### Regulation Compliance under the ASIC Act, the Life Insurance Act 1995 and APRA's prudential standards. #### Deadlines Claims and litigation work has tight deadlines, jurisdiction-specific procedure and AFCA or court filing duties. Errors cost money or cause prejudice. #### Insurers' own AI Advice on insurers' AI in underwriting, claims triage and fraud detection engages CPS 230, ICA s.13 and Privacy Act APP 1.7 automated-decision disclosure (from 10 December 2026). ### New Zealand #### Disputes No AFCA. Disputes go to industry dispute-resolution schemes and the courts. #### Tribunal The Canterbury Earthquakes Insurance Tribunal is listed in the Courts of New Zealand GenAI guidelines. #### Injury claims Check any personal-injury component against ACC cover. Sources: - [Courts of New Zealand: GenAI guidelines — Lawyers (7 December 2023)](https://www.courtsofnz.govt.nz/assets/6-Going-to-Court/practice-directions/practice-guidelines/all-benches/20231207-GenAI-Guidelines-Lawyers.pdf) (checked 2026-09-23) - [Accident Compensation Act 2001](https://www.legislation.govt.nz/act/public/2001/49/en/latest/) (checked 2026-09-23) - [Community Law: Overview of the ACC scheme](https://communitylaw.org.nz/community-law-manual/chapter-19-accident-compensation-acc/overview-of-the-acc-scheme/) (checked 2026-09-23) ### United States Local name: Insurance law #### Governing law Insurance regulation is mainly state (McCarran-Ferguson), with some federal overlays. Policy interpretation is state contract and insurance-code law. #### Checking Formal Opinion 512: verify coverage citations and wording against the actual policy. Sources: - [ABA Formal Opinion 512, Generative Artificial Intelligence Tools (29 July 2024)](https://www.americanbar.org/content/dam/aba/administrative/professional_responsibility/ethics-opinions/aba-formal-opinion-512.pdf) (checked 2026-09-23) - [15 U.S.C. § 1011 (McCarran-Ferguson Act): declaration of policy](https://www.law.cornell.edu/uscode/text/15/1011) (checked 2026-09-23) - [15 U.S.C. § 1012: regulation by State law](https://www.law.cornell.edu/uscode/text/15/1012) (checked 2026-09-23) *Practice guidance for planning and review. Confirm current primary sources and the requirements of the actual matter. This is general information, not legal advice.*