Submission 2978 — AUSCL Australasian Society for Computers & Law — NDIS Future Generations Bill

No machine-readable text was extracted from this page.

Submission to the Senate Community Affairs

Legislation Committee

National Disability Insurance Scheme Amendment

(Securing the NDIS for Future Generations) Bill 2026

4 July 2026

AUSCL

The Australasian Society for Computers and Law (AUSCL) is a multidisciplinary think tank and educational charity concerned with the intersection of law, technology and society. Its members include lawyers, technologists, academics, policy specialists, public-sector professionals, privacy and cybersecurity experts, legal operations specialists, computational law practitioners and AI governance professionals.

AUSCL contributes to public policy on artificial intelligence, privacy, cybersecurity, digital identity, automated decision-making, legal technology, data governance, online safety and digital regulation.

Executive position AUSCL welcomes the opportunity to provide this submission to the Senate Community Affairs Legislation Committee in relation to the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026 (Bill).

AUSCL supports digital transformation of the National Disability Insurance Scheme (NDIS), including technology that improves timeliness, consistency, accessibility, integrity and lawful administration. In the context of this Bill, that transformation must be responsible and safe because it concerns high-impact decisions affecting people with disability, including access, eligibility, supports, funding, plan administration, compliance action, debt, suspension and review pathways.

The NDIS is not merely a funding scheme; it is a statutory framework that reflects Australia’s commitments to the rights, inclusion and social participation of people with disability. Decisions under the Bill engage dignity, autonomy, equality, privacy, health, independent living, social participation, access to justice and the right to an effective remedy.

AUSCL endorses the Human Technology Institute’s (HTI) submission, particularly its analysis that the Bill’s automation provisions are facilitative rather than protective: they authorise automated decision-making (ADM) in the NDIS without sufficient statutory boundaries around what is permissible, impermissible and reviewable. HTI identifies three necessary conditions for legitimate ADM: a robust legal and governance framework, reliable human-and-technology decision-making, and accessible accountability and appeal protections. [3]

AUSCL also supports the Law Council of Australia’s concerns about legislative process, rule-of-law safeguards, review rights, broad Ministerial powers and the automation of discretionary or evaluative decision-making. [4]

AUSCL’s contribution adds a privacy, human rights, AI governance, assurance and data-protection perspective. The central recommendation is that the Bill be amended before passage to ensure that any use of ADM or data-driven decision support in the NDIS is lawful, rights-preserving, transparent, contestable, independently assured and subject to accountable human responsibility.

Executive summary AUSCL observes that the Bill enables high-impact, data-driven public administration in the NDIS, a scheme affecting the dignity, autonomy, equality, privacy, health, independent living and social participation of people with disability. This requires a clear statutory framework for legality, transparency, contestability, human responsibility, independent assurance and parliamentary scrutiny.

Automation is not a mere machinery matter. In the NDIS context, it may determine or materially influence access to essential supports. The rule of law requires clear legal authority, identifiable human responsibility, intelligible reasons, procedural fairness, effective review, data protection and independent oversight.

AUSCL does not oppose automation in principle. In clear, objective and beneficial contexts, it can improve timeliness and consistency. However, automation is inappropriate where it removes individualised assessment, obscures legal responsibility, narrows review rights, intensifies surveillance, repurposes sensitive disability and health information, or scales unlawful or unfair outcomes.

AUSCL also recommends consideration of regulatory digital twins and related computational law tools to improve legislative quality and pre-enactment impact assessment. These tools can model proposed NDIS amendments against existing law, policy objectives, participant scenarios, human rights obligations, review pathways, data flows and operational constraints before enactment.

This would assist Parliament to identify unintended consequences, compare less rights-restrictive drafting options, assess distributional and access-to-justice impacts, and create a verifiable evidence trail for scrutiny. AUSCL would welcome the opportunity to assist in assessing the feasibility of a regulatory digital twin sandbox or pilot.

Key recommendations AUSCL recommends that the Bill be amended to:

Recommendation Amendment

Recommendation 1 – Limited Prohibit full automation of discretion, evaluative judgment and automated discretion state-of-mind formation where the action may reduce, suspend, limit or otherwise affect participant rights, entitlements, supports, review pathways or human rights

Recommendation 2 – Primary Confine automation authorisation to categories specified in legislation limits primary legislation, with clear identification of the responsible human office-holder and applicable safeguards

Recommendation 3 – Remove or substantially constrain proposed s 59C(2), Constrained Ministerial including by requiring disallowability, consultation and an powers express bar on expanding automation into discretionary or evaluative decisions by legislative instrument

Recommendation 4 – Preserve internal review, external merits review and judicial Preserved review rights review for any decision made by, or materially influenced by,

ADM

Recommendation 5 – Require accessible notice, meaningful reasons, disclosure of Accessible notice and the material role of ADM, correction pathways and a right to reasons human reconsideration

Recommendation 6 – Impact Require privacy impact assessments and human rights impact assessments required assessments before deployment or material change of any

NDIS ADM system

Recommendation 7 – Public Establish a public NDIS ADM register ADM register

Recommendation 8 – Mandate independent assurance before and after deployment, Independent ADM assurance including data quality, accuracy, bias, disparate-impact, cybersecurity, accessibility and legal-compliance testing

Recommendation 9 – Make human responsibility operational, not nominal, by Meaningful human identifying a senior human decision owner and ensuring staff responsibility have training, authority and time to depart from system outputs

Recommendation 10 – Embed privacy-by-design and data minimisation, including Privacy-by-design controls limits on secondary use, retention, linkage, profiling and reuse of sensitive disability, health, behavioural, financial and location-related information

Recommendation 11 – High- Treat functional capacity, support-needs, budget-setting, impact ADM tools planning, record-keeping, debt and compliance tools as high impact ADM where their outputs materially influence participant outcomes

Recommendation 12 – Prevent Ensure that record-keeping, compliance and debt provisions data-driven hardship do not create automated or data-driven hardship for participants who did not control relevant records, relied on providers or nominees, or faced disability-related, safety related, digital-access, language, housing or system barriers

Recommendation 13 – Assess the feasibility of a regulatory digital twin sandbox or Regulatory digital twin pilot pilot to test proposed NDIS amendments against existing law, policy objectives, participant scenarios, human rights obligations, review pathways and operational impacts before enactment

Analysis

For the Committee’s ease of reference, Annexure A provides a Clause-specific amendment map. Set out below is further analysis of the following issues and recommendations:

A. Stronger rule-of-law safeguards B. ADM in high-impact public administration C. Automation of discretionary and evaluative judgment D. Deeming automated decisions to be CEO decisions is insufficient E. Ministerial automation powers removed or constrained F. Functional capacity and planning tools as high-impact ADM G. Privacy and data governance safeguards are essential H. Avoiding data-driven hardship in compliance and debt processes I. Practical and enforceable contestability J. Meaningful human oversight

K. Independent assurance and public reporting L. ADM register M. Scrutiny of Ministerial support and budget powers N. Justice impact assessment and legal assistance O. UN SDG alignment P. Adoption of regulatory digital twins for evidence-based legislative reform A. Stronger rule-of-law safeguards

Recommendation

AUSCL recommends that the Bill not proceed in its current form. At minimum, it should be amended to ensure clear legal authority, procedural fairness, proportionality, human rights compatibility, accessible contestability, merits review, independent oversight, accountable human decision making and genuine consultation with people with disability and their representative organisations

The Bill concerns a statutory scheme that provides essential supports to hundreds of thousands of Australians with disability. It warrants close scrutiny against fundamental rule-of-law requirements, including clear legal authority, certainty, procedural fairness, reviewability, proportionality, human rights compatibility and parliamentary accountability.

The Law Council has cautioned that rushed reform in this context can undermine meaningful engagement with affected communities and limit early identification of unintended consequences.

[4] National Legal Aid, Community Legal Centres Australia, the Law Council and NATSILS have also

warned that the reforms are significant and complex, will affect people’s daily lives and futures, and may substantially increase legal need unless accessible review pathways and legal assistance are strengthened. [5]

The Convention on the Rights of Persons with Disabilities requires States Parties to closely consult with, and actively involve, people with disability, including through their representative organisations, in decisions that concern them. [12] This is a substantive human rights obligation directly relevant to legislative changes affecting access, supports, planning, review and administrative decision-making.

AUSCL recommends that the Bill not proceed in its current form. At minimum, it requires amendment to preserve individualised assessment, accessible contestability, merits review, independent oversight and accountable human decision-making.

B. ADM in high-impact public administration

Recommendation

AUSCL recommends that the treat all automated or data-driven tools that materially influence NDIS outcomes as high-impact ADM, whether or not formally labelled as automated decision-making. Such systems should be subject to clear legal authority, administrative law compliance, privacy and human rights safeguards, independent assurance, accessible reasons, human reconsideration and effective review rights.

The Parliamentary Library Bills Digest describes the Bill as part of a broader package directed to scheme sustainability, fraud control, provider regulation and governance. [2] Efficiency, sustainability and fraud prevention are legitimate public objectives. They do not displace administrative law, privacy law, human rights or the right to effective review.

The Commonwealth Ombudsman’s updated ADM Better Practice Guide states that automated systems can make decisions, recommend decisions, provide decision support, generate preliminary assessments and automate fact-finding. It also states that automated systems should comply with administrative law values, including legality, fairness, rationality and transparency. [6]

The Office of Parliamentary Counsel’s ADM guidance similarly identifies risks including algorithmic bias, inaccurate or less accurate decisions, unclear reasons, large volumes of inaccurate decisions made quickly and the risk that inflexible automated criteria may fetter discretion. [7]

These risks are acute in the NDIS. Disability-related support needs are contextual, dynamic and often intersect with health, housing, income, geography, culture, family violence, caring responsibilities, language, digital access and other social determinants. ADM systems that reduce individual circumstances to fixed categories, scores or historical patterns can produce superficially consistent but substantively unfair outcomes.

The automation risk is not confined to systems formally labelled as ADM. Functional capacity tools, support-needs assessments, budget-setting methods, standardised planning tools, record-keeping systems, compliance analytics, debt analytics, fraud detection tools and Ministerial instruments can each materially influence participant outcomes. AUSCL recommends that the Committee assess the Bill as creating or enabling a connected data-driven decision system, not merely a set of discrete automation powers.

C. Automation of discretionary and evaluative judgment

Recommendation

AUSCL recommends that the Bill prohibit computer programs from making, materially influencing or determining administrative actions involving discretion, evaluative judgment or formation of a state of mind where the action may reduce, suspend, limit or otherwise affect participant rights, entitlements, supports, review pathways or human rights, except for narrowly defined, objective and clearly beneficial actions.

The most serious concern with Schedule 3, Part 2, Division 5 is that proposed s 59B appears to authorise automation involving discretion, evaluative judgment and formation of a state of mind. HTI describes this as a radical feature of the Bill and potentially unprecedented in Australian law. [3]

Administrative discretion is not an inconvenience to be engineered away. It is a legal mechanism for doing justice in individual cases. It allows decision-makers to consider context, exceptionality, vulnerability, competing evidence, proportionality and statutory purpose. In the NDIS context, discretion and evaluative judgment are often essential to fair decision-making.

Automation is structurally ill-suited to many discretionary tasks because automated systems tend to standardise, classify, infer and optimise. They can flatten individual circumstances, reproduce biased assumptions, obscure reasons, create false precision and encourage human decision makers to over-rely on system outputs.

The NSW Ombudsman has warned that if ADM in a discretionary decision-making system means that the administrator is no longer able, or does not in practice, continue to exercise genuine discretion, the system will be inconsistent with the statute that granted the discretion and its outputs will be unlawful. [3]

AUSCL recommends amending the Bill to prohibit a computer program from making, materially influencing or determining administrative action involving discretion, evaluative judgment or formation of a state of mind where the action may reduce, suspend, limit or otherwise affect participant rights, entitlements, supports, review pathways or human rights, except for narrowly defined, objective and clearly beneficial actions.

D. Deeming automated decisions to be CEO decisions is insufficient

Recommendation

Deeming automated actions to be CEO decisions may assist accountability, but further safeguards are needed to ensure each automated or materially automated decision is legally cognisable, attributable to an identified responsible officer, supported by intelligible reasons, transparent as to the role of ADM, capable of human reconsideration, and subject to internal, external merits and judicial review.

The Bill appears to deem certain automated actions to have been taken by the CEO. That deeming provision may be intended to preserve accountability and review rights. However, deeming a computer-generated action to be taken by the CEO does not, by itself, ensure that there has been a legally cognisable decision, meaningful human consideration, intelligible reasons, procedural fairness or effective review.

The Full Federal Court’s decision in Pintarich v Deputy Commissioner of Taxation [2018] FCAFC 79; (2018) 262 FCR 41 (Full court of the Federal Court of Australia, Kerr, Moshinsky and Derrington JJ, 25 May 2018) illustrates the difficulty of applying traditional administrative law concepts to automated correspondence and automated outputs. The case raises questions about when an authorised

officer has actually made a decision and whether traditional conceptions of a “decision” remain adequate in the digital era. [19]

The Law Council correctly identifies this as a serious judicial review issue. [4] If the law does not clearly identify the decision-maker, decision, reasons, data relied upon and review pathway, affected participants may be left with formal rights that are practically ineffective.

The Bill should expressly provide that every automated or materially automated decision:

• is a decision for the purposes of internal review, external merits review and judicial review;

• is taken to have been made by an identified legally responsible officer;

• must be accompanied by reasons intelligible to the affected participant;

• must disclose the use and material role of ADM;

• must identify the data, rule, model output, score, assessment result or compliance flag materially relied upon;

• must be capable of human reconsideration; and

• must not limit any review right that would have existed had the decision been made by a human decision-maker.

E. Ministerial automation powers removed or constrained

Recommendation

AUSCL recommends the proposed s 59C(2) be removed or tightly limited so that automation cannot be expanded by legislative instrument into high-impact, discretionary or rights-affecting NDIS decisions. Any retained power should be disallowable and subject to consultation, privacy and human rights impact assessment, public explanation, independent assurance and sunset review.

AUSCL supports HTI’s recommendation that proposed s 59C(2) be removed or substantially constrained. [3] Decisions about whether NDIS decisions may be automated are not merely technical implementation matters. They are questions of legal authority, public accountability, human rights, privacy and democratic legitimacy.

The Minister should not be able to expand automation by legislative instrument into high-impact decisions involving access, eligibility, plan content, support levels, compliance, debt, suspension, review rights, discretion or evaluative judgment.

If any such power is retained, it should be subject to the following statutory limits:

• express prohibition on specifying provisions involving discretion, evaluative judgment or formation of a state of mind;

• disallowability;

• mandatory consultation with people with disability and disability representative organisations;

• privacy impact assessment;

• human rights impact assessment;

• anti-discrimination and disparate-impact assessment;

• public explanation of the legal basis, decision class, safeguards and review rights;

• independent assurance before commencement; and
• sunset review. Administrative instruments and standard operating procedures have a role, but they cannot carry the burden of core rights protection. The central safeguards belong in primary legislation.

F. Functional capacity and planning tools as high-impact ADM

Recommendation

Functional capacity, support-needs, planning, budget-setting and renewal tools should be treated as high-impact ADM where they materially influence participant outcomes. Such tools should be co designed, cohort-validated, accessibility-tested, independently audited, supported by plain language explanations and subject to human reconsideration before any adverse outcome takes effect.

The Bill’s automation risk is broader than proposed ss 59B-59E. Functional capacity assessments, support-needs assessments, planning tools, budget methods and automated renewal processes may materially shape access, eligibility, funding, plan content, reassessment and review outcomes.

AUSCL recommends that any tool that materially influences participant outcomes be treated as high-impact ADM, even if a human officer formally remains involved.

Such tools require:

• co-design with people with disability and representative organisations;

• validation for relevant cohorts;

• accessibility testing;

• testing for fluctuating, progressive and complex disability;

• testing for communication disability, psychosocial disability, acquired brain injury and neurodivergence;

• testing for First Nations participants, culturally and linguistically diverse participants and participants in regional or remote communities;

• publication of plain-language information about how the tool works;

• accessible reasons where the tool materially affects an outcome;

• human reconsideration before any adverse outcome takes effect; and

• independent audit and periodic reporting. The Federal Court’s decisions in National Disability Insurance Agency v Davis [2022] FCA 1002 (Federal Court of Australia, Mortimer J on 29 August 2022) (Davis) and Chief Executive Officer of the National Disability Insurance Agency v Eastham [2026] FCA 147 (Federal Court of Australia, Hespe J on 25 February 2026) (Eastham) demonstrate the importance of contextual and whole-person assessment in NDIS decision-making.

In Davis, the Federal Court of Australia considered permanence and the meaning of available treatment in the context of access rules. [20] In Eastham, the Court confirmed that the relevant impairment need not be the sole cause of the need for a support; it is sufficient if it is a cause of the

need for the support. [21] The Bill should not enable standardised or automated assessment architecture to reverse these principles in practice.

G. Privacy and data governance safeguards are essential

Recommendation

AUSCL recommends that each NDIS ADM system be subject to privacy impact assessment before deployment and material change, addressing legal authority, data minimisation, purpose limitation, sensitive information handling, data quality, profiling, secondary use, vendor access, security, retention, correction rights, audit logging and safeguards for participants relying on nominees, advocates, carers or supported decision-making.

NDIS ADM is likely to use or infer sensitive information, including disability, health, support needs, functional capacity, service use, family and nominee information, location, financial information, provider interactions and behavioural indicators.

Privacy risk arises not only at collection. It also arises through linkage, inference, profiling, retention, secondary use, vendor access, model training, audit logging and downstream reuse.

From 10 December 2026, APP entities that arrange for a computer program to use personal information to make decisions that could reasonably be expected to significantly affect an individual’s rights or interests will be required to include additional information in their APP privacy policies. [8] The OAIC has also consulted on guidance for transparency in ADM, noting that the Privacy Act amendments introduce an ADM transparency obligation from that date. [9]

Those reforms support, but do not replace, stronger NDIS-specific safeguards in primary legislation. Privacy policy transparency alone is not sufficient for a high-impact disability support scheme.

AUSCL recommends that each NDIS ADM system be subject to a privacy impact assessment before deployment and before material change. The OAIC describes a privacy impact assessment as a systematic assessment of a project’s privacy impacts and recommendations for managing, minimising or eliminating those impacts. [10]

For NDIS ADM, privacy impact assessment should address:

• legal authority for collection, use and disclosure;
• data minimisation;
• purpose limitation;
• data quality and provenance;
• consent, authority and participant comprehension;
• sensitive information handling;
• security and access controls;
• retention and deletion;
• correction rights;
• audit logging;

• vendor and subcontractor access;
• model training prohibitions or controls;
• linkage and profiling risks;
• secondary use and function creep;
• cross-system reuse in compliance, debt, fraud or enforcement contexts; and
• safeguards for people who rely on nominees, advocates, carers or supported decision-making. H. Avoiding data-driven hardship in compliance and debt processes

Recommendation

AUSCL recommends that no debt, penalty, suspension or adverse compliance action proceed from missing, inconsistent or inaccessible records without clear legal authority, accessible notice, reasons, human verification, reasonable excuse and hardship exceptions, opportunity to correct records, review rights, and safeguards against treating missing records as an automated proxy for fraud or non-compliance.

Schedule 2, Part 4, including proposed s 45B(5), would impose record-retention obligations on participants who make claims for NDIS amounts. Where record-keeping obligations are unclear or onerous, missing or inconsistent records may become adverse risk signals in debt, compliance or fraud analytics. That is inappropriate without strong statutory safeguards.

Participants may not control the relevant records. They may rely on providers, nominees, plan managers or informal supports. They may also face disability-related, safety-related, digital-access, language, housing, remoteness or provider-dependence barriers.

AUSCL recommends that the Bill provide that no debt, penalty, suspension or adverse compliance action arising from missing, inconsistent or inaccessible records can proceed without:

• clear legal authority;
• accessible notice;
• reasons;
• human verification;
• reasonable excuse and hardship exceptions;
• consideration of disability-related and safety-related barriers;
• opportunity to correct records;
• access to relevant information;
• review rights; and
• prohibition on treating missing records as an automated proxy for fraud or non-compliance. Sensitive disability and health information collected for support should not be repurposed into profiling or enforcement systems without transparent statutory controls, necessity, proportionality, data minimisation and independent assurance.

I. Practical and enforceable contestability

Recommendation

AUSCL recommends that ADM notice and review rights be accessible, timely and enforceable, including plain-language disclosure of whether ADM was used, how it materially influenced the outcome, what facts, data and criteria were relied on, who is accountable, how information can be corrected, and how human reconsideration, internal review, external merits review and legal assistance can be accessed.

The Australian Human Rights Commission’s Human Rights and Technology Final Report

recommends human rights impact assessment for AI systems used by government in administrative decision-making and emphasises transparency and accountability for AI-informed decisions. [11]

For the NDIS, relevant rights include equality and non-discrimination, access to justice, privacy, social security, health, independent living, participation in community life, dignity, autonomy and freedom from arbitrary or disproportionate interference.

Contestability cannot be satisfied by a formal legal right that is inaccessible in practice. Participants need to understand:

• whether ADM has been used;

• whether ADM materially influenced the outcome;

• what material facts, data, rules, scores, assessment results or compliance flags were relied upon;

• what legal criteria were applied;

• who is accountable for the decision;

• how to correct incorrect information;

• how to seek human reconsideration;

• how to obtain internal review;

• how to obtain external merits review; and

• where to obtain advocacy or legal assistance. Notice under proposed s 59E warrants strengthening. Notice needs to be accessible, timely, in plain language, available in the participant’s preferred communication mode, and capable of being understood by a participant, nominee, advocate or supporter. Review periods are best calculated from the point at which valid accessible notice has been given, with legal consequences where proper notice is not provided.

J. Meaningful human oversight

Recommendation

AUSCL recommends that each ADM system have a named senior human decision owner, clear legal responsibility, trained staff with authority and time to depart from system outputs, audit logs of human acceptance or departure, and vendor contracts that preserve audit, access, remediation, data handling and system-change rights.

Human responsibility must be operational, not symbolic.

A human reviewer needs authority, competence, time, information and institutional permission to depart from an automated output. Human oversight is only meaningful where staff can understand the basis of a system recommendation, access reasons, question the output, and depart from it without operational or institutional pressure to treat the automated result as presumptively correct.

AUSCL recommends that the Bill require:

• a named senior human decision owner for each ADM system;

• clear legal responsibility for each decision class;

• trained frontline staff with authority to depart from system outputs;

• no requirement for exceptional escalation before departure from an automated output;

• staff training in administrative law, human rights, disability rights, privacy and automation risk;

• audit logs showing when humans accepted, queried or departed from system outputs;

• vendor contracts with audit, access, logging, remediation, data handling and system-change rights; and

• consultation with staff before deployment of systems they are required to use, rely on or review. The Digital Transformation Agency’s AI technical standard provides practical guidance for Australian Government agencies embedding AI in government systems and states that the standard sets consistent practices across the AI lifecycle. [16] AUSCL recommends that NDIS automation be subject to at least that level of lifecycle governance, with additional safeguards because of the high impact disability context.

K. Independent assurance and public reporting

Recommendation

AUSCL recommends that NDIS ADM systems be independently assured before deployment and periodically thereafter, with testing for legal authority, administrative law compliance, privacy, human rights, data quality, accuracy, reliability, subgroup error rates, bias, cybersecurity, accessibility, vendor controls, auditability, explanation quality, complaints, review outcomes and error remediation.

Robodebt demonstrated that unlawful or unsafe automated decision-making can persist when systems are opaque, responsibility is diffuse, legal concerns are not escalated and affected individuals cannot effectively challenge decisions. The Royal Commission into the Robodebt Scheme recommended legislative reform to introduce a consistent legal framework for automation in government services. [13]

The Commonwealth Ombudsman’s 2025 report on the Targeted Compliance Framework found serious problems where automated systems continued to operate inconsistently with law and policy after legislative change. [14] The ANAO’s report on administration of the Age Pension illustrates the

scale of harm that can arise where high-volume social security administration produces underpayments and overpayments. [15]

The NDIS is precisely the kind of high-impact setting where small error rates can produce large aggregate harm. Faster processing is not a sufficient public-interest justification where automation can scale unlawful, inaccurate, inaccessible or discriminatory outcomes.

AUSCL recommends independent assurance before deployment and at regular intervals throughout the system lifecycle. Assurance would test:

• legal authority;
• administrative law compliance;
• privacy compliance;
• human rights compatibility;
• data quality and provenance;
• accuracy and reliability;
• subgroup error rates;
• bias and disparate impact;
• cybersecurity;
• accessibility;
• system drift;
• vendor controls;
• audit logging;
• explanation quality;
• complaints;
• review outcomes; and
• remediation of errors. NIST’s AI Risk Management Framework identifies trustworthy AI characteristics including validity and reliability, safety, security and resilience, accountability and transparency, explainability and interpretability, privacy enhancement and fairness with harmful bias managed. [17] Those characteristics warrant translation into enforceable statutory obligations for NDIS ADM.

L. ADM register

Recommendation

AUSCL recommends a statutory NDIS ADM register which identifies each system’s purpose, legal authority, affected decisions and cohorts, data sources, model or rules category, vendor involvement, human oversight, responsible officer, impact assessments, assurance results, known limitations, complaints, review outcomes, error rates, material changes and systemic remediation.

AUSCL recommends a statutory NDIS ADM register. The register could identify, for each system:

• system name and purpose;
• legal authority;

• decision or administrative action affected;
• whether the system makes, recommends, supports or materially influences decisions;
• affected participant cohorts;
• data sources;
• model, rules engine or system category;
• vendor involvement;
• human oversight arrangements;
• responsible senior officer;
• privacy impact assessment status;
• human rights impact assessment status;
• assurance results;
• accuracy and known limitations;
• complaint data;
• review outcomes;
• error rates;
• material changes; and
• incidents or systemic remediation. Public transparency can be balanced with genuinely necessary security and system-integrity protections. Opacity, however, is best treated as the exception rather than the default.

M. Scrutiny of Ministerial support and budget powers

Recommendation

AUSCL recommends that any Ministerial instrument affecting NDIS support entitlements or participant rights be disallowable, subject to consultation, human rights compatibility assessment, privacy and data impact assessment where relevant, differential impact analysis, merits review where individual entitlements are affected, and a necessity, justification and proportionality requirement.

The Law Council has raised serious concerns about broad Ministerial powers to affect support categories, old framework plans, indexation, maximum funding, support intensity and plan renewal by legislative instrument. [4]

AUSCL shares those concerns. Formulae, instruments and budget methods can materially reduce supports without a practical, individualised review pathway. They may operate as automation adjacent decision architecture even where not formally described as ADM.

AUSCL recommends any Ministerial instrument affecting support entitlements or participant rights be:

• disallowable;
• subject to consultation with people with disability and representative organisations;
• accompanied by a statement of compatibility with human rights;

• accompanied by a privacy and data impact assessment where data-driven methods are used;

• assessed for differential impact on First Nations participants, women and girls, culturally and linguistically diverse participants, people in regional and remote areas, people with psychosocial disability and people with complex support needs;

• subject to merits review where it affects individual entitlements; and

• limited by an express statutory requirement that the measure be necessary, justified, proportionate and the least rights-restrictive option reasonably available.

N. Justice impact assessment and legal assistance

Recommendation

AUSCL recommends that the Bill be accompanied by a justice impact assessment estimating impacts on internal review, ART applications, courts, legal assistance, disability advocacy, regional and remote participants, participants with communication or cognitive barriers, First Nations and culturally diverse participants, and the costs of unlawful, erroneous or poorly explained decisions.

The Bill is likely to increase internal reviews, complaints, Administrative Review Tribunal applications and court proceedings. The Law Council has recommended justice impact assessment as a mechanism for assessing the impact of legislative and regulatory change on the justice system. [22]

AUSCL recommends that the Bill be accompanied by a justice impact assessment estimating:

• expected increases in internal review;
• expected increases in ART applications;
• expected increases in Federal Court and Federal Circuit and Family Court matters;
• legal assistance and advocacy funding required;
• impacts on rural, regional and remote participants;
• impacts on participants with cognitive, psychosocial, communication or literacy barriers;
• impacts on First Nations participants and culturally and linguistically diverse participants; and
• costs of unlawful, erroneous or poorly explained decisions. AUSCL further recommends that the Bill not commence until legal assistance, disability advocacy, review and oversight systems are adequately funded.

O. UN SDG alignment

Recommendation

AUSCL recommends that NDIS ADM reforms be assessed against relevant UN Sustainable Development Goals, including health and wellbeing, gender equality, reduced inequalities, and peace, justice and strong institutions, to ensure automation does not undermine access to supports, increase unpaid care burdens, entrench disparate impacts or weaken transparency, reviewability and accountability.

The AUSCL supports the UN Sustainable Development Goals (UN SDG). The UN SDG provide a useful public-policy lens for the Bill. The United Nations identifies 17 Sustainable Development Goals, including good health and wellbeing, gender equality, reduced inequalities, and peace, justice and strong institutions. [18]

In this context:

• SDG 3 - Good health and wellbeing: NDIS ADM must not delay, reduce or distort access to disability supports in ways that undermine health, safety or independent living.

• SDG 5 - Gender equality: Changes to eligibility, supports and contestability may have gendered effects, including additional unpaid care burdens for women and girls.

• SDG 10 - Reduced inequalities: ADM needs safeguards against disparate impact on people with disability, First Nations people, people in regional and remote communities, culturally and linguistically diverse communities, and people with limited digital access.

• SDG 16 - Peace, justice and strong institutions: NDIS automation requires transparency, reviewability, accessible justice, public accountability and independent oversight.

P. Regulatory digital twins for evidence-based legislative reform

Recommendation

AUSCL recommends that significant NDIS reforms be accompanied by a computational impact assessment using a regulatory digital twin to model proposed amendments against existing law, policy objectives, participant scenarios, human rights, privacy, review pathways and operational impacts before enactment. AUSCL would welcome the opportunity to assist in assessing a transparent, rights-preserving and independently assured sandbox or pilot.

AUSCL supports the modernisation of legislation through regulatory digital twins. AUSCL recommends that the Australian Government adopt regulatory digital twins and related computational law tools to improve the quality, transparency and impact assessment of NDIS legislative reform before enactment.

A regulatory digital twin is a machine-readable, testable and version-controlled representation of the legal, policy, operational and data logic of a regulatory scheme. Properly designed, it can represent legislation, rules, policy settings, decision pathways, eligibility criteria, obligations, review rights, data flows, operational constraints and implementation logic. It allows proposed amendments to be modelled against existing law, intended policy objectives and real-world participant scenarios before legislation is passed.

For complex schemes such as the NDIS, this capability is increasingly important. The Bill affects eligibility, supports, funding, planning, compliance, record-keeping, automation, debt and review pathways. These elements interact dynamically. A textual amendment that appears minor may materially alter participant outcomes, administrative burdens, review rights, privacy exposure, downstream costs to other systems, or the practical availability of remedies. Traditional explanatory memoranda and manual legal review remain essential, but they are not always sufficient to detect these effects at scale or to test alternative legislative designs before enactment.

AUSCL recommends the use of a regulatory digital twin before major NDIS amendments are introduced or passed, to:

• model the legal effect of proposed amendments against the current NDIS Act, Rules, operational guidance and review pathways;

• test proposed amendments against policy objectives, including sustainability, fraud prevention, dignity, autonomy, equality, accessibility, individualised support, privacy, procedural fairness and reviewability;

• simulate impacts on representative participant cohorts, including people with psychosocial disability, cognitive disability, communication disability, acquired brain injury, fluctuating conditions, First Nations participants, culturally and linguistically diverse participants, regional and remote participants, children, women and girls with disability, and participants requiring supported decision-making;

• identify unintended consequences, including loss of eligibility, reduced supports, cost- shifting to families or other service systems, increased debt risk, increased review applications, or reduced practical access to remedies;

• assess compatibility with human rights and administrative law principles, including the CRPD, privacy law, procedural fairness, merits review, non-discrimination and proportionality;

• compare alternative drafting options, allowing Parliament to consider less rights-restrictive or more administratively workable amendments before enactment;

• generate traceable amendment impact statements showing how each proposed change affects rights, obligations, powers, discretions, review pathways, data flows and affected cohorts; and

• support post-enactment monitoring by comparing predicted impacts with actual operational data, complaints, review outcomes, error rates and participant experience.

This approach is not speculative. In Australia, Realta Logic describes its platform as a “Digital Twin for Compliance” and “Digital Regulatory Infrastructure” designed to operationalise regulation in context across organisations.[23]

OpenFisca provides an open-source rules-as-code engine for collaboratively modelling laws and regulations and making them computable through open APIs; its documentation explains that it can describe tax and benefit systems and calculate outcomes from factual situations.[24] France has used OpenFisca-based approaches for open calculation models and microsimulation of tax and social benefit systems; the Open Government Partnership records France’s OpenFisca commitment as an open API enabling microsimulation of the French tax and social benefit system.[25] OpenFisca materials also record that France’s National Assembly created LexImpact to enable Members of Parliament to evaluate reforms on test cases and across the French population within seconds.[26] New Zealand’s Better Rules work began in 2018 as a collaboration involving Inland Revenue, the Ministry of Business, Innovation and Employment and the Parliamentary Counsel Office to explore machine-readable legislation-as-code, while OpenFisca Aotearoa provides an open-source project for computational models of New Zealand legislation, regulation and government policy.[27]

AUSCL further recommends that blockchain or other tamper-evident ledger technologies be considered where appropriate to support integrity, provenance and auditability of regulatory digital twin artefacts. This does not require sensitive participant data to be stored on-chain. Rather, cryptographic hashes, verifiable credentials or other tamper-evident proofs could be used to verify the integrity of:

• versions of draft amendments;
• policy assumptions;
• computational rules;
• test scenarios;
• impact assessments;
• consultation inputs;
• assurance reports; and
• post-implementation review evidence. Used properly, blockchain-enabled provenance can help ensure that Parliament, oversight bodies and affected communities can verify which version of a legislative model was tested, what assumptions were applied, what evidence was considered, and whether later changes departed from the assessed model.

These tools are best designed to support, not replace, democratic law-making. Regulatory digital twins and verifiable audit trails can complement public consultation, parliamentary scrutiny, human rights assessment, privacy impact assessment and expert legal analysis. They are not a substitute for transparent, accountable and participatory legislative deliberation, and require safeguards against becoming opaque automated policy-making systems.

AUSCL recommends that the Committee propose a statutory or administrative requirement that significant NDIS reforms be accompanied by a computational impact assessment, including:

• a machine-readable model of the proposed amendments;
• comparison with the existing legislative framework;
• testing against policy, human rights, privacy and administrative law objectives;
• scenario-based impact analysis for affected participant cohorts;
• identification of unintended consequences and less restrictive alternatives;
• independent assurance of the computational model and assumptions;
• publication of a plain-language summary; and
• a verifiable evidence trail sufficient for parliamentary, public and expert scrutiny. This would improve legislative quality, reduce implementation risk, strengthen parliamentary scrutiny, and help ensure that NDIS reform is not only fiscally sustainable, but legally coherent, rights-compatible, operationally workable and trusted by participants.

AUSCL would welcome the opportunity to assist the Committee, the Department and relevant oversight bodies in exploring the feasibility of a regulatory digital twin sandbox or pilot for future NDIS reform. AUSCL’s Future Law Working Group includes expertise in computational law, rules-as

code, digital regulatory infrastructure, privacy, AI governance, legal technology, administrative law and human-centred system design.

A carefully scoped pilot could test how machine-readable legislative models, participant scenarios, policy objectives, rights safeguards, review pathways, data-flow mapping and assurance artefacts could be used to assess proposed amendments before enactment.

Such a sandbox would need to be transparent, rights-preserving, independently assured and developed with people with disability and their representative organisations. It could support Parliament and government to identify unintended consequences, compare alternative drafting options, strengthen explanatory materials and improve implementation readiness, while preserving democratic deliberation, legal judgment and public consultation.

Conclusion

AUSCL supports careful and rights-preserving use of technology to improve NDIS administration. It does not oppose automation in principle. In clear, objective and beneficial contexts, automation can improve timeliness and consistency.

However, the Bill currently authorises high-impact ADM and data-driven administration without adequate statutory boundaries, privacy protections, human rights safeguards, accessible contestability, independent assurance or accountable human responsibility.

AUSCL endorses HTI’s submission, supports the Law Council’s rule-of-law concerns, and recommends amendment of the Bill before passage. Responsible NDIS automation requires more than legal authorisation for computerised administration. It requires careful scrutiny of the Bill’s wider data-driven operating architecture, including functional capacity assessment, support-needs assessment, planning, budget-setting, record-keeping, debt, compliance analytics, accessible notice, cohort validation and review pathways.

AUSCL recommends that the Committee recommend that the Bill not proceed in its current form. At minimum, the Bill warrants amendment to ensure the NDIS remains lawful, transparent, accessible, independently assured, privacy-preserving, human-rights compatible and trusted by the people whose lives it affects.

Annexure A Clause-specific amendment map

Provision / issue Proposed amendment or safeguard

Schedule 1, Part 1 - Functional capacity assessment would be strengthened by taking proposed s 9B - into account the person’s real-world environment, communication Functional capacity needs, assistive technology, supports, safety, fluctuating conditions, cultural context, service availability and supported decision-making needs. Assessment tools that materially influence access, eligibility or funding warrant co-design, cohort-specific validation, accessible reasons, human reconsideration and independent assurance.

Schedule 2, Part 4 - Record-keeping failure is better treated as a contextual issue rather proposed s 45B(5) and than an automatic debt or compliance signal. The legislation would proposed s 182(4) - be improved by recognising reasonable excuse, hardship, safety, Records and debt provider-dependence, nominee-dependence, digital exclusion, language and disability-related barriers, with human verification and review before any debt, penalty or suspension action.

Schedule 3, Part 2, Automation of discretion, evaluative judgment or state-of-mind Division 5 - proposed ss formation should be prevented where the action may reduce, 59B(3)-(5) - Automation of suspend, limit or otherwise affect participant rights, entitlements, administrative action supports, review pathways or human rights, except for narrowly defined, objective and clearly beneficial actions.

Proposed s 59B (5) - Deeming action to be taken by the CEO is insufficient. The Bill Deeming CEO action would benefit from identifying the legally responsible human office holder, requiring intelligible reasons, preserving review rights, and ensuring a trained human reviewer has authority to depart from the automated output.

Proposed s 59B (11) Review rights are central to any decision made by, or materially Schedule 3, Part 2, item influenced by, ADM, including internal review, external merits review 12 and judicial review. Provisions that limit substitution of a correct or Proposed s 202 (2A) preferable decision warrant deletion or revision to preserve effective accountability and participant remedies.

Proposed s 59C (2) - The power is best removed or tightly confined. Any retained power Ministerial specification would require disallowability, consultation with disability of further automation representative organisations, and an express prohibition on extending automation to broad discretion, evaluative judgment, rights-affecting decisions, support reductions, debt, compliance action, suspension or review pathways.

Provision / issue Proposed amendment or safeguard

Proposed s 59D - Any standard operating procedure would benefit from statutory Standard operating minimum content, including: legal authority, eligible decision class, procedures data inputs, excluded uses, human oversight, accessibility controls, privacy and human rights assessment, assurance, audit logging, error escalation, change control, vendor controls, public reporting and retirement criteria.

Proposed s 59E - Notice of Notice requirements warrant enforceable legal consequences. automation Participants need accessible notice, reasons, disclosure of the material role of ADM, correction pathways, human reconsideration and review rights. Review periods are best calculated only from the point at which valid accessible notice has been given.

Ministerial support Provisions affecting support categories, old framework plans, determinations, plan indexation, maximum funding, automatic renewal or support renewal and budget intensity warrant assessment for automation-adjacent impact. methods Formulae, instruments and budget methods can materially reduce supports without a practical individualised review pathway and are best accompanied by disallowance, consultation, human rights assessment and review safeguards.

ADM register and The Bill would be strengthened by requiring a public ADM register independent assurance and independent pre-deployment and periodic assurance for each high-impact ADM or decision-support system used in the NDIS.

Privacy and data The Bill would be strengthened by requiring privacy-by-design, data governance minimisation, and clear limits on secondary use, retention, linkage, profiling and reuse of sensitive disability, health, behavioural, financial and location-related information.

References

[1] Senate Community Affairs Legislation Committee, National Disability Insurance Scheme

Amendment (Securing the NDIS for Future Generations) Bill 2026 inquiry page. https://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Community_Affairs/NDI SFutureGenBill

[2] Parliamentary Library, Bills Digest No. 65, 2025-26, National Disability Insurance Scheme

Amendment (Securing the NDIS for Future Generations) Bill 2026. https://www.aph.gov.au/Parliamentary_Business/Bills_Legislation/bd/bd2526/26bd065

[3] Human Technology Institute, NDIS Amendment (Securing the NDIS for Future Generations) Bill

2026 submission page. https://www.uts.edu.au/research/centres/human-technology institute/submissions

[4] Law Council of Australia, Inquiry into National Disability Insurance Scheme Amendment

(Safeguarding the NDIS for Future Generations) Bill 2026. https://lawcouncil.au/resources/submissions/inquiry-into-national-disability-insurance scheme-amendment-safeguarding-the-ndis-for-future-generations-bill-2026

[5] National Legal Aid, Community Legal Centres Australia, Law Council of Australia and NATSILS,

Joint statement: NDIS reforms risk real harm without proper consultation (2 June 2026). https://nationallegalaid.org.au/news/joint-statement-ndis-reforms-risk; Joint statement PDF. https://nla-production-assets.s3.ap-southeast 2.amazonaws.com/public/Statements/2026/Joint-statement-NDIS-reforms-risk-real-harm without-proper-consultation.pdf

[6] Commonwealth Ombudsman, Automated Decision-Making: Better Practice Guide (March 2025). https://www.ombudsman.gov.au/__data/assets/pdf_file/0025/317437/Automated-Decision

Making-Better-Practice-Guide-March-2025.pdf

[7] Office of Parliamentary Counsel, Instructor Guidance Note - Automated decision-making (2026). https://www.opc.gov.au/sites/default/files/2026-03/Instructor%20Guidance%20Note%20 %20Automated%20decision-making.docx; Automated decision-making (2025 PDF version): https://www.opc.gov.au/sites/default/files/2025-11/Automated%20decision-making.pdf

[8] OAIC, Chapter 1: APP 1 - Open and transparent management of personal information. https://www.oaic.gov.au/privacy/australian-privacy-principles/australian-privacy-principles guidelines/chapter-1-app-1-open-and-transparent-management-of-personal-information

[9] OAIC, Consultation on guidance for transparency in automated decision-making. https://www.oaic.gov.au/engage-with-us/consultations/consultation-on-guidance-for transparency-in-automated-decision-making

[10] OAIC, Privacy impact assessment tool. https://www.oaic.gov.au/privacy/privacy-guidance-for organisations-and-government-agencies/privacy-impact-assessments/privacy-impact assessment-tool

[11] Australian Human Rights Commission, Human Rights and Technology: Final Report (2021). https://humanrights.gov.au/our-work/technology-and-human-rights/publications/final-report human-rights-and-technology; PDF version: https://humanrights.gov.au/sites/default/files/document/publication/ahrc_rightstech_2021_fin al_report_10.pdf

[12] OHCHR, Convention on the Rights of Persons with Disabilities. https://www.ohchr.org/en/instruments-mechanisms/instruments/convention-rights-persons disabilities

[13] Royal Commission into the Robodebt Scheme, Final Report (2023). https://robodebt.royalcommission.gov.au/publications/report

[14] Commonwealth Ombudsman, Automation in the Targeted Compliance Framework: when the law is changed but the system isn’t (August 2025). https://www.ombudsman.gov.au/__data/assets/pdf_file/0017/320750/Automation-in-the

Targeted-Compliance-Framework.pdf

[15] Australian National Audit Office, Administration of the Age Pension, Auditor-General Report No. 20 of 2025-26. https://www.anao.gov.au/work/performance-audit/administration-of-the-age pension

[15] Australian National Audit Office, Administration of the Age Pension (PDF). https://www.anao.gov.au/sites/default/files/2026-01/Auditor-General_Report_2025-26_20.pdf

[16] Digital Transformation Agency, Technical standard for government’s use of artificial intelligence. https://www.digital.gov.au/policy/ai/AI-technical-standard

[16] Digital Transformation Agency, AI technical standard: introduction, scope and target audience. https://www.digital.gov.au/policy/ai/AI-technical-standard/technical-standard-governments use-artificial-intelligence-introduction-scope-and-target-audience

[17] NIST, Artificial Intelligence Risk Management Framework (AI RMF 1.0).

https://www.nist.gov/itl/ai-risk-management-framework; NIST, AI Risk Management Framework

1.0 PDF. https://nvlpubs.nist.gov/nistpubs/ai/nist.ai.100-1.pdf; NIST AI RMF, trustworthy AI characteristics. https://airc.nist.gov/airmf-resources/airmf/3-sec-characteristics/

[18] United Nations, Sustainable Development Goals. https://sdgs.un.org/goals

[19] Pintarich v Deputy Commissioner of Taxation [2018] FCAFC 79. https://jade.io/summary/mnc/2018/FCAFC/79

[20] National Disability Insurance Agency v Davis [2022] FCA 1002. https://jade.io/summary/mnc/2022/FCA/1002

[21] Chief Executive Officer of the National Disability Insurance Agency v Eastham [2026] FCA 147. https://jade.io/summary/mnc/2026/FCA/147

[22] Law Council of Australia, Policy Statement - Justice Impact Assessments (12 September 2013). https://lawcouncil.au/publicassets/cf70dcfc-bdd6-e611-80d2-005056be66b1/1309-Policy

Statement-Justice-Impact-Assessment-Policy-Statement.pdf

[23] Realta Logic, LinkedIn company profile, describing Realta Logic as “Digital Twin for Compliance” and “Digital Regulatory Infrastructure” designed to operationalise regulation across organisations: https://au.linkedin.com/company/realta-logic. See also Stone & Chalk, ‘No more red tape: How Realta Logic helps teams move faster’ (29 November 2023), describing Realta Logic’s use of digital twins as virtual models of company operations directly linked to real regulations: https://www.stoneandchalk.com.au/articles/meet-our-founders-siobhaine-slevin.

[24] OpenFisca, ‘Write rules as code’, describing OpenFisca as enabling economists and researchers to simulate the impact of past or future reforms on income distribution: https://openfisca.org/en/. See also OpenFisca Documentation, ‘Tax and Benefit System’, stating that the OpenFisca core engine can simulate a country’s legislation once represented as source code: https://openfisca.org/doc/key-concepts/tax_and_benefit_system.html

[25] Open Government Partnership, France, ‘Open Calculation Models and Simulators (FR0019)’, describing OpenFisca as an open search engine and open API allowing microsimulation of the French tax and social benefit system: https://www.opengovpartnership.org/members/france/commitments/FR0019/.

[26] OpenFisca, ‘About’, recording that the first LexImpact simulator was created by the French National Assembly in 2019 to enable Members of Parliament to evaluate reforms on test cases and the whole French population within seconds: https://openfisca.org/en/about/. See also Laboratoire Société Numérique, ‘OpenFisca: when a digital commons turns law into code’ (12 December 2023), describing LexImpact as enabling MPs to simulate the effects of amendments and assess impacts on typical cases, public finances and income redistribution: https://labo.societenumerique.gouv.fr/en/articles/openfisca-quand-un-commun-numerique transforme-la-loi-en-code/.

[27] New Zealand Service Innovation Lab, ‘Better Rules and Legislation as Code’, recording that the Better Rules work began in early 2018 as a collaboration with Inland Revenue, the Ministry of Business, Innovation and Employment, the Parliamentary Counsel Office and a private software company to explore machine-readable legislation-as-code: https://serviceinnovationlab.github.io/projects/legislation-as-code/. See also New Zealand Government, ‘Better Rules for Government Discovery Report’, describing an approach to policy development and implementation that creates a common, multidisciplinary understanding of rules across the whole process: https://www.digital.govt.nz/dmsdocument/95-better-rules-for government-discovery-report/html; GitHub, ‘BetterRules/openfisca-aotearoa’, open-source repository for computational models of New Zealand legislation, regulation and government policy: https://github.com/BetterRules/openfisca-aotearoa

Contributors

AUSCL Future Law Working Group

AUSCL Sustainability Working Group

Contact Details

contact@auscl.org

Attention: AUSCL Lead Policy Analyst

Disclaimer

This submission has been prepared by the above Contributors for the purpose of contributing to public policy, law reform and regulatory discussions. The views and recommendations expressed in this submission do not necessarily reflect the views of any individual member, contributor, employer, client, government agency, academic institution or other organisation with which a contributor may be associated. This submission is intended to provide general information and policy analysis only. It does not constitute legal advice, professional advice or any other form of advice and should not be relied upon as such. Readers should obtain independent professional advice appropriate to their circumstances. This submission has been informed by the expertise of the Contributors’ members and working groups across law, technology, government, academia and industry. Participation in the development of this submission does not imply endorsement of every view, recommendation or conclusion contained within it. While reasonable care has been taken in the preparation of this submission, the Contributors make no representation or warranty as to the completeness, accuracy or currency of the information contained in it and accepts no liability for any loss or damage arising from reliance upon its contents. Copyright © Australasian Society for Computers and Law (AUSCL). This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited.

Suggested citation Australasian Society for Computers and Law (2026). Submission to the Senate Community Affairs Legislation Committee, National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill. AUSCL Policy Series PS-2026-13