1. Introduction

Digital therapeutics (DTx) refers to evidence-based therapeutic interventions delivered through software, mobile applications, artificial intelligence (AI), connected medical devices, and digital platforms to prevent, manage, or treat medical conditions.

Unlike ordinary health and wellness applications, digital therapeutics are designed to produce measurable clinical benefits through therapeutic interventions.

Examples include:

Software-based cognitive behavioural therapy for depression and anxiety.

Digital interventions for diabetes management.

Smartphone applications supporting substance-use disorder treatment.

AI-assisted rehabilitation and neurological therapy.

Software-based interventions for chronic pain.

Digital programmes for sleep disorders and cardiovascular risk management.

In Australia, digital therapeutics regulation primarily involves the Therapeutic Goods Administration (TGA), the Therapeutic Goods Act 1989 (Cth), the Therapeutic Goods Regulations 1990 (Cth), medical-device classification rules, privacy legislation, and clinical evidence requirements.

The central legal question is whether a digital therapeutic is merely a general wellness product or a regulated therapeutic good or medical device.

This distinction determines its regulatory obligations, including registration, clinical evidence, safety monitoring, advertising restrictions, and legal liability.

2. Legislative framework governing digital therapeutics

The principal legislation and regulatory instruments include:

Section 41BD of the Therapeutic Goods Act 1989 is particularly important. It establishes the statutory definition of a medical device, including software intended for specified medical purposes.

The TGA's digital therapeutics guidance, updated on 30 January 2026, expressly identifies DTx as a category of Software as a Medical Device (SaMD).

Therapeutic Goods Administration (TGA)+1

3. Classification of digital therapeutics

Digital therapeutics are not automatically classified as low-risk software. Their classification depends on intended purpose, functionality and potential harm.

The TGA uses four principal medical-device classes:

These examples are illustrative rather than automatic classifications. The applicable classification rules must be assessed for each product.

Important legal principle: A developer cannot avoid medical-device regulation merely by describing a therapeutic application as a lifestyle or wellness product. The actual intended purpose, representations, functionality and instructions for use are important.

Therapeutic Goods Administration (TGA)+1

4. TGA approval and Australian Register of Therapeutic Goods (ARTG)

The ARTG is the central regulatory register for therapeutic goods legally supplied in Australia.

Where a digital therapeutic qualifies as a medical device and is neither excluded nor exempt, it must be included in the ARTG before supply.

The regulatory process generally involves:

Identifying the product's intended medical purpose.

Determining whether it falls within the statutory definition of a medical device.

Assessing whether an exclusion or exemption applies.

Determining its appropriate risk classification.

Establishing the applicable conformity-assessment requirements.

Preparing technical and clinical documentation.

Obtaining the necessary conformity-assessment evidence.

Applying for ARTG inclusion through the Australian regulatory sponsor.

Maintaining compliance after market entry.

An overseas developer intending to supply a DTx product in Australia must also consider the Australian sponsor requirements.

Legal significance: ARTG inclusion is not simply a commercial registration. It is part of Australia's statutory system for controlling the safety, quality and performance of therapeutic goods.

Therapeutic Goods Administration (TGA)+1

5. Clinical evidence and scientific validation

Clinical evidence is one of the most important aspects of digital therapeutics regulation.

A digital therapeutic making claims about treating depression, diabetes, chronic pain or neurological disorders must have appropriate scientific support for its intended medical purpose.

The evidence may include:

Randomised controlled clinical trials.

Clinical effectiveness studies.

Real-world performance data.

Clinical validation studies.

Usability and human-factors assessments.

Software verification and validation.

Evidence concerning adverse effects and patient safety.

For example, suppose an Australian biotechnology company develops an AI-based application claiming to reduce symptoms of major depressive disorder.

The company should not rely solely on user satisfaction surveys or the number of application downloads. It needs appropriate clinical evidence supporting the therapeutic claims.

Clinical evidence should also address the relevant patient population, treatment duration, clinical endpoints, limitations and foreseeable risks.

Regulatory distinction: Evidence of technical functionality is not necessarily evidence of clinical effectiveness. A software application can operate exactly as programmed and still fail to produce the claimed therapeutic benefit.

Therapeutic Goods Administration (TGA)+1

6. AI-based digital therapeutics and algorithmic accountability

Artificial intelligence creates additional regulatory challenges.

Some digital therapeutics use machine-learning algorithms to:

Recommend treatment programmes.

Monitor patient symptoms.

Predict deterioration.

Personalise therapeutic interventions.

Analyse behavioural and physiological information.

Modify treatment recommendations based on patient responses.

The legal problem arises when an algorithm produces an inaccurate or unsafe therapeutic recommendation.

Developers and sponsors must consider:

A. Algorithmic transparency: Appropriate documentation of the algorithm's purpose, performance and limitations.

B. Data quality: Whether training and validation datasets adequately represent the intended patient population.

C. Bias and discrimination: Whether the system produces systematically less accurate results for particular patient groups.

D. Software updates: Whether changes introduce new medical functions or alter the device's risk classification.

E. Human oversight: Whether the product requires professional supervision and whether users understand its limitations.

A software update can have regulatory consequences. For example, a wellness application that introduces automated disease diagnosis or treatment recommendations may become a regulated medical device.

The TGA's software guidance specifically recognises that changes in functionality and intended purpose can affect regulatory status and classification.

Therapeutic Goods Administration (TGA)+1

7. Digital therapeutics and patient privacy

Digital therapeutics frequently collect sensitive personal information, including:

Psychiatric and psychological records.

Genetic and genomic information.

Medication histories.

Biometric measurements.

Sleep and behavioural information.

Physiological monitoring data.

The Privacy Act 1988 (Cth) and Australian Privacy Principles (APPs) are particularly relevant to organisations covered by the Act.

Important privacy obligations include:

Collecting personal information only where legally permitted.

Providing appropriate privacy notices.

Obtaining consent where required.

Maintaining appropriate information security.

Limiting secondary use and disclosure.

Providing access and correction rights.

Managing eligible data breaches under the Notifiable Data Breaches scheme.

Cross-border data transfers create additional concerns under APP 8, which regulates disclosure of personal information to overseas recipients.

Example: An Australian digital mental-health company transfers patient records to an overseas cloud service provider. The company must consider whether the overseas disclosure complies with APP 8 and the associated accountability requirements.

8. Advertising and consumer protection

Digital therapeutic manufacturers must ensure that advertising and promotional statements comply with applicable therapeutic-goods advertising requirements and the Australian Consumer Law.

Potentially unlawful conduct may include:

Claiming an application cures cancer without adequate evidence.

Advertising an unapproved therapeutic product as TGA-approved.

Exaggerating the clinical effectiveness of an AI treatment platform.

Concealing significant limitations or risks.

Making unsupported claims about replacing professional medical treatment.

Section 18 of the Australian Consumer Law prohibits misleading or deceptive conduct in trade or commerce.

Section 29 addresses false or misleading representations concerning goods and services.

A developer may therefore face separate regulatory and consumer-law consequences for misleading therapeutic claims.

9. Exemptions and exclusions for digital therapeutics

Not every software product with a health-related function must be included in the ARTG.

Australian regulation recognises two important concepts.

A. Excluded software

Certain products are excluded under the Therapeutic Goods (Excluded Goods) Determination 2018.

Excluded software does not require ARTG inclusion under the TGA framework.

B. Exempt software

Some medical-device software may be exempt from ARTG inclusion while remaining subject to specified regulatory obligations.

For example, certain clinical decision support software may qualify for exemption if it satisfies all applicable statutory criteria.

However, patient-facing digital therapeutics cannot automatically rely on the exemption available to certain professional-facing clinical decision support systems.

A product that directly provides treatment recommendations or therapeutic interventions to patients may require ARTG inclusion.

The TGA guidance distinguishes excluded products from exempt products and explains that exempt devices can remain subject to safety, advertising and adverse-event reporting obligations.

Therapeutic Goods Administration (TGA)+2

10. Post-market surveillance and adverse-event liability

Regulation does not end when a digital therapeutic enters the Australian market.

Manufacturers and sponsors must maintain appropriate systems for identifying, investigating and responding to safety problems.

Important areas include:

Monitoring adverse events and product complaints.

Investigating software malfunctions.

Maintaining cybersecurity protections.

Correcting clinically significant software errors.

Communicating safety warnings.

Implementing recalls where necessary.

Assessing the impact of software modifications.

Maintaining appropriate technical and clinical documentation.

Example:

An AI-based digital therapeutic recommends an inappropriate insulin-management intervention because of a software error.

The error causes a patient to suffer serious injury.

Potential legal consequences may involve:

TGA investigation and regulatory enforcement.

Product liability under the Australian Consumer Law.

Negligence claims against responsible parties.

Contractual liability.

Privacy or cybersecurity consequences if personal information is also compromised.

The precise liability depends on causation, the parties' responsibilities, applicable statutory provisions and the available evidence.

11. Case laws and judicial principles

Important distinction

There is an important limitation in Australian digital therapeutics jurisprudence: the leading reported decisions discussed below concern medical treatment, medical devices and consumer protection generally. They are not judgments specifically deciding the ARTG classification of a digital therapeutic application.

Nevertheless, they provide important legal principles for interpreting DTx liability and governance.

Case 1. Rogers v Whitaker (1992) 175 CLR 479

High Court of Australia

Facts: The case concerned medical treatment and the failure to adequately warn a patient about a material risk associated with an operation.

Legal issue: Whether a medical practitioner had a duty to disclose material risks of proposed treatment.

Judgment: The High Court recognised that a risk is material if a reasonable person in the patient's position would likely attach significance to it, or if the practitioner knew or ought to have known that the particular patient would attach significance to it.

Relevance to digital therapeutics:

This principle is relevant when digital therapeutic products are used as part of clinical treatment.

For example, a patient using an AI-powered mental-health application should receive appropriate information about:

Known therapeutic limitations.

Potential risks.

The possibility of inaccurate recommendations.

The role of professional supervision.

Circumstances requiring conventional medical treatment.

The case supports the importance of informed decision-making and adequate disclosure in digital healthcare.

Case 2. Ethicon Sàrl v Gill [2021] FCAFC 29

Full Court of the Federal Court of Australia

Facts: The litigation involved defective urogynaecological medical devices and allegations concerning inadequate safety evaluations, warnings and misleading representations.

Legal issues:

Whether the devices were defective.

Whether consumers were entitled to expect a higher level of safety.

Whether adequate pre-market and post-market evaluation had been undertaken.

Whether warnings and marketing representations were misleading.

Whether the manufacturers had breached their duty of care.

Judicial significance: The case examined product safety, consumer expectations, manufacturer responsibility, warnings, causation and negligence in the context of medical devices.

Relevance to digital therapeutics:

The same broad principles are important when a software-based therapeutic causes patient harm.

For example, if a manufacturer releases a digital therapeutic with an inadequately tested algorithm, fails to communicate known risks, or makes misleading safety claims, questions may arise under product liability and negligence principles.

The case is particularly relevant to the responsibility of manufacturers to undertake adequate safety evaluations throughout a medical device's lifecycle.

CaseNote AU

Case 3. Australian Competition and Consumer Commission v TPG Internet Pty Ltd [2013] HCA 54

High Court of Australia

Facts: The dispute concerned advertising for telecommunications services and whether prominent advertising claims were misleading despite qualifications appearing elsewhere.

Legal issue: Whether the overall impression of an advertisement could mislead consumers.

Judicial significance: The High Court considered the advertisement as a whole, including the prominence of the headline representation and the qualifications accompanying it.

Relevance to digital therapeutics:

A digital therapeutic company cannot necessarily correct a misleading headline claim by placing important qualifications in small print.

For example:

An application advertises that it eliminates depression.

The limitations are disclosed only in lengthy terms and conditions.

Consumers are encouraged to believe that the application replaces professional treatment.

Such promotional practices may raise concerns under section 18 of the Australian Consumer Law.

The case provides a useful interpretive principle for evaluating digital-health advertising.

Case 4. Grant v Australian Knitting Mills Ltd [1936] AC 85

Privy Council appeal from Australia

Facts: The case concerned injury caused by defective clothing containing harmful chemical residue.

Legal issue: Whether a manufacturer owed a duty of care to the ultimate consumer despite the absence of a direct contractual relationship.

Judicial significance: The decision is an important foundation of Australian product liability and manufacturer negligence principles.

Relevance to digital therapeutics:

A patient injured by a defective therapeutic application may have a potential negligence claim against a manufacturer even without a direct contractual relationship, subject to the applicable legal requirements.

The principles are relevant to software design defects, inadequate testing and foreseeable patient harm.

Case 5. ACCC v Reckitt Benckiser (Australia) Pty Ltd [2016] FCAFC 181

Full Court of the Federal Court of Australia

Facts: The case involved representations concerning the effectiveness of Nurofen products for specific types of pain.

Legal issue: Whether product representations created misleading impressions about the products' effectiveness and specificity.

Judicial significance: The proceedings illustrate the importance of accurate product representations and the prohibition against misleading therapeutic marketing.

Relevance to digital therapeutics:

A DTx manufacturer must ensure that its marketing accurately reflects the clinical evidence.

For example, an application should not claim that its treatment programme is clinically superior to conventional therapy without adequate supporting evidence.

The principles are relevant to claims concerning therapeutic effectiveness, personalised treatment and comparative clinical performance.

12. Biotech industry-specific regulatory challenges

Digital therapeutics increasingly overlap with biotechnology, genomics, pharmaceutical research and personalised medicine.

This creates several additional legal issues.

A. Digital therapeutics and genomic medicine

Applications may analyse genetic information to personalise treatment programmes.

Legal issues include:

Protection of genomic information.

Accuracy of genetic predictions.

Discrimination risks.

Consent for secondary use of genomic data.

Responsibility for incorrect treatment recommendations.

B. Digital therapeutics and pharmaceutical integration

Some digital therapeutics operate alongside prescription medicines.

For example, a diabetes application may support medication adherence and monitor glucose measurements.

Legal questions include:

Whether the software is a medical device.

Whether its therapeutic claims are adequately supported.

Whether it provides medical advice beyond its intended purpose.

Whether its integration with another medical device changes its classification.

C. Digital therapeutics and biotechnology clinical trials

Digital interventions may be incorporated into clinical trials involving new medicines, vaccines or biological products.

The relevant legal considerations include:

Research ethics approval.

Informed consent.

Clinical-trial data integrity.

Patient confidentiality.

Software validation.

Reporting of adverse events.

The National Statement on Ethical Conduct in Human Research and applicable research governance requirements are important in this area.

D. Digital therapeutics and reproductive biotechnology

Software used to monitor fertility, support conception or guide reproductive interventions may fall within the medical-device framework depending on its intended purpose.

Such products can also raise privacy, reproductive autonomy and discrimination concerns.

13. Liability of manufacturers, developers and healthcare providers

Australian law does not place every digital-health responsibility on a single participant.

A particularly difficult issue is determining liability when several parties contribute to a defective therapeutic outcome.

For example, an overseas developer creates an AI treatment algorithm, an Australian sponsor markets it, and a hospital incorporates it into clinical treatment.

If the algorithm generates harmful advice, the legal investigation may need to examine the developer's software design, sponsor compliance, clinical supervision and hospital implementation.

Liability must be determined according to the particular conduct, applicable legal duties and causation evidence.

14. Regulatory reforms and emerging developments in 2026

Australia's digital-health regulatory framework is evolving.

Important developments include:

The TGA's dedicated digital therapeutics guidance, updated on 30 January 2026.

Updated guidance on software-based medical devices published on 24 February 2026.

Updated clinical decision support software guidance dated 16 October 2026.

Continued regulatory attention to AI-based medical devices.

Review of the exclusion for certain digital mental-health tools, with consultation open until 7 December 2026.

The last development is particularly significant because digital mental-health applications can operate between general wellness services and regulated therapeutic interventions.

The TGA's regulatory review aims to examine whether the existing treatment of certain digital mental-health products remains appropriate. The consultation should not be mistaken for an already enacted change in the law.

Therapeutic Goods Administration (TGA)+3

15. Constitutional and administrative law dimensions

Digital therapeutics regulation also raises questions concerning the exercise of governmental regulatory powers.

The TGA operates within the Commonwealth statutory framework. Its decisions may affect:

Market access for biotechnology companies.

Availability of innovative medical treatments.

Patient safety and public health.

Regulatory fairness.

Commercial freedom and innovation.

Potential administrative-law disputes include:

Unreasonable regulatory decisions.

Procedural fairness in regulatory decision-making.

Incorrect application of medical-device classification rules.

Disputes concerning ARTG inclusion or cancellation.

Regulatory enforcement and judicial review.

The Administrative Decisions (Judicial Review) Act 1977 (Cth) may be relevant where a reviewable decision falls within its scope. Other judicial review remedies may also be available under applicable law.

16. Critical legal problems and future challenges

17. Suggestions for strengthening Australia's digital therapeutics framework

Several measures could improve regulatory accountability:

1. Clearer classification rules: Develop more detailed guidance for distinguishing wellness applications from therapeutic software.

2. Stronger clinical validation: Require evidence proportionate to the therapeutic risks and intended clinical benefits.

3. AI accountability: Establish appropriate documentation, algorithmic monitoring and change-management requirements.

4. Patient protection: Improve disclosure of therapeutic limitations, risks and data practices.

5. Post-market monitoring: Strengthen mechanisms for detecting software failures and adverse clinical outcomes.

6. Cross-border cooperation: Improve coordination between the TGA and overseas medical-device regulators.

7. Research governance: Strengthen ethical oversight of digital therapeutics used in biotechnology clinical trials.

8. Clear liability allocation: Improve contractual and regulatory clarity concerning manufacturers, sponsors, developers and healthcare providers.

18. Conclusion

Digital Therapeutics Regulation in Australia represents an important development in biotechnology and medical-device law.

The Therapeutic Goods Act 1989 (Cth), section 41BD, and the Therapeutic Goods (Medical Devices) Regulations 2002 (Cth) provide the principal regulatory foundation.

The TGA treats digital therapeutics as Software as a Medical Device when they satisfy the statutory definition. Such products generally require ARTG inclusion unless an applicable exclusion or exemption applies.

The cases Rogers v Whitaker, Ethicon Sàrl v Gill, ACCC v TPG Internet Pty Ltd, Grant v Australian Knitting Mills Ltd and ACCC v Reckitt Benckiser (Australia) Pty Ltd provide important principles concerning informed consent, medical-device safety, manufacturer negligence and misleading therapeutic representations.

Although these decisions do not directly establish a separate digital-therapeutics legal doctrine, they offer useful foundations for resolving disputes involving software-based medical interventions.

Ultimately, Australia's regulatory challenge is to balance biotechnology innovation, patient autonomy, clinical effectiveness, algorithmic accountability and public health protection.

19. Important legislation and research references

For further academic and legal research, the following primary sources are particularly useful:

Therapeutic Goods Act 1989 (Cth).

Therapeutic Goods (Medical Devices) Regulations 2002.

TGA – Digital Therapeutics.

TGA – Regulation of Software-Based Medical Devices.

Privacy Act 1988 (Cth).

Competition and Consumer Act 2010 (Cth).

TGA – Advertising Software-Based Medical Devices.

Academic research note: The case laws above are general medical negligence, medical-device and consumer-protection authorities applied by analogy. They should not be represented as direct judicial decisions on digital therapeutics regulation.

I can also prepare the next topic, Digital Therapeutics Clinical Trial Liability and Governance in Australia – Detailed Explanation With Case Laws, in the same legal-research format.