Focus modes, sleep trackers, brainwave apps: when does health software need TGA approval?

Whether a brainwave app or sleep tool needs TGA approval depends on one thing: what the manufacturer claims it does. This post explains the SaMD classification line, what Class IIa and IIb mean in practice, and the question every health software founder should answer before launch.

Abstract diagram showing layered geometric wave forms suggesting frequency bands and regulatory classification hierarchy, in navy and cream
25 Jun 2026
4
 min read

Whether a brainwave app, sleep tool, or cognitive focus product is classified as Software as a Medical Device (SaMD) in Australia comes down to one thing: intended purpose. The TGA does not regulate based on what a product tracks or measures. It regulates based on what the manufacturer says the product does. An app that helps you wind down is a wellness tool. The same app claiming to treat insomnia or reduce ADHD symptoms is a medical device and needs ARTG inclusion before it can be legally supplied.

MIT researchers have spent the past decade studying what happens when you expose the brain to 40 Hz light and sound, a gamma frequency. Phase II clinical trials from MIT spinoff Cognito Therapeutics showed that people with Alzheimer's exposed to 40 Hz sensory stimulation experienced significant slowing of brain atrophy and improvements on cognitive measures. A 2023 study in Scotland using audio and visual stimulation at 37.5 Hz found improvements in memory recall across over 100 participants. A Phase III trial is now underway.

Consumer products have followed the science. EEG headbands, focus headphones, binaural beat apps, and cognitive state monitors are now mainstream, promising to shift how your brain operates. Australia is paying attention: BrainTech, a conference for the local neurotechnology ecosystem, held its inaugural event in 2025 and returns in 2026. The Australian Government invested A$135 million in digital mental health in early 2025. The TGA has made it a priority.

And that is where the regulatory question enters the picture.

The science behind brainwave entrainment

The brain generates electrical activity in rhythmic patterns, measurable by EEG, across frequency bands that shift depending on what you are doing or thinking. Delta waves (0.5-4 Hz) dominate in deep sleep. Theta waves (4-8 Hz) are linked to creativity and the drowsy state between waking and sleep. Alpha waves (8-12 Hz) characterise relaxed alertness. Beta waves (13-30 Hz) reflect active focus. Gamma waves (30-100 Hz) are associated with peak cognition, memory binding, and high-level information processing.

Brain entrainment (also called audio-visual entrainment, or AVE) uses rhythmic sensory stimuli to guide the brain toward a target frequency state. The mechanism draws on the brain's tendency to synchronise its oscillations with external rhythmic input. Apply a 40 Hz audio tone or flickering light, and the cortex shifts toward 40 Hz activity. The consumer pitch follows naturally: entrain to beta for focus, theta for sleep, gamma for cognitive performance.

One important caveat. The MIT gamma research is primarily Alzheimer's-focused. The clinical trials are studying reduction of amyloid pathology, not general productivity gains. The consumer claim that gamma entrainment improves everyday focus is plausible but not what the trials are testing.

The TGA's one question: what does the manufacturer say it does?

The TGA's position, restated in its February 2026 guidance on software medical devices, is consistent. The regulatory trigger is intended purpose, not technology.

It does not matter whether a product uses EEG sensors, binaural beats, near-infrared spectroscopy, or 40 Hz light pulses. What matters is what the manufacturer says the product is for.

Under the Therapeutic Goods Act, software is a medical device if it is intended to diagnose, prevent, monitor, treat, or alleviate a disease or condition. The moment a brainwave product's marketing, labelling, or website steps into that territory, ARTG inclusion is required before it can be legally supplied in Australia.

The wellness side of the line: "supports relaxation," "helps you wind down," "designed for focus." Consumer claims governed by the ACCC.

The SaMD side: "clinically proven to reduce anxiety," "treats insomnia," "assists with ADHD symptom management." Those claims invoke treatment of a named condition. ARTG territory.

The same app. Different claims. Entirely different regulatory obligations.

What SaMD classification actually requires

If a brainwave or cognitive software product is classified as SaMD, three things must happen before it can be supplied in Australia.

Register on the ARTG. The Australian Register of Therapeutic Goods is the gateway to legal supply. Required evidence depends on device class. Class IIa SaMD (moderate risk, non-critical monitoring) requires a conformity assessment and a technical file including clinical evaluation. Class IIb, covering higher-risk treatment-decision support software, requires a more demanding evidence package and additional scrutiny.

Build a clinical evidence base. This means a structured clinical evaluation that meets the TGA's framework, not marketing references to university research. Where the commercial roadmap includes the US, the same evidence base needs to be designed to support an FDA 510(k) or De Novo submission from the start. FDA standards on study design and data volume are typically more demanding than TGA requirements. Building the evidence once, for both regulators, costs significantly less than retrofitting a technical file after launch.

Establish post-market surveillance. SaMD manufacturers must monitor real-world device performance and report adverse events on an ongoing basis. This is not a one-time compliance task. It is a continuous obligation tied to the ARTG registration.

A well-constructed ARTG registration, designed with a dual-pathway strategy in mind, is a commercial asset. It signals to investors and acquirers that the product's efficacy story has been stress-tested by a regulator, not just cited in a pitch deck.

The question every health software founder should answer before launch

Does your product diagnose, monitor, treat, prevent, or alleviate a specific disease or condition?

If the honest answer is no, stay in the wellness lane. Build consumer claims that are substantiated under the ACL, sharp enough to be compelling, and defensible if the ACCC asks.

If the answer is yes, or might be yes within 12 months, engage the SaMD pathway early. The classification position, and the evidence built around it, becomes foundational to everything downstream: investor materials, partnership terms, market access.

Intended purpose is something you control. The TGA responds to what manufacturers claim. Founders who make that call deliberately, before launch, are in a considerably stronger position than those who discover the answer mid-enforcement.

Frequently asked questions

What is brain entrainment and how does it relate to SaMD?

Brain entrainment uses rhythmic stimuli (sound frequencies like binaural beats and isochronic tones, or light pulses) to guide the brain toward a target frequency state. Whether products qualify as Software as a Medical Device depends not on the technology, but on the intended medical purpose the manufacturer claims. The same EEG headband can be a wellness product or a regulated medical device depending entirely on how it is described and promoted.

How does the TGA decide if a health app is a medical device?

The TGA applies the intended purpose test. If software is designed to diagnose, monitor, treat, prevent, or alleviate a disease or condition, it is a medical device under the Therapeutic Goods Act and needs ARTG inclusion before it can be supplied in Australia. The platform (app, web, wearable) and the underlying technology are irrelevant to this test. What the company says the product does, across marketing, labelling, and instructions, is what the TGA looks at.

What is the difference between a wellness app and a SaMD in Australia?

A wellness app makes general health claims and sits under Australian Consumer Law, regulated by the ACCC. A SaMD makes specific medical claims, diagnosing, treating, or monitoring a named disease or condition, and sits under the Therapeutic Goods Act, regulated by the TGA. The same underlying technology can fall into either category depending entirely on how the product is positioned and promoted.

What are SaMD Class IIa and Class IIb in Australia?

These are TGA risk classifications for Software as a Medical Device. Class IIa covers moderate-risk software, typically products that monitor or inform clinical decisions without directly driving treatment. Class IIb covers higher-risk software where incorrect output could lead to serious harm, such as treatment-decision support tools. Class IIb requires a more substantial evidence package and greater regulatory scrutiny than Class IIa. The classification determines the conformity assessment pathway and the depth of clinical evidence required for ARTG inclusion.

What changed with the TGA's February 2026 SaMD guidance?

The TGA confirmed that AI-enabled and software-based medical devices, including real-time neural monitoring tools, are regulated under the existing SaMD framework, not a separate technology-specific regime. The guidance reinforced the intended-purpose test: technology type is irrelevant. If a product claims a medical purpose, it is a medical device. The TGA also clarified manufacturer liability for companies deploying third-party software tools with medical functions.

What happens if a company makes medical claims without ARTG registration?

Supplying a product that meets the definition of a medical device without ARTG registration is a breach of the Therapeutic Goods Act, carrying significant civil and criminal penalties. The TGA has signalled an active enforcement posture on digital health tools making medical claims without registration, following its October 2025 review of digital mental health tools. Beyond TGA exposure, unregistered medical claims also create ACCC liability under the Australian Consumer Law's misleading conduct provisions.

How Parallaxis helps

Classification aside, the harder work is building evidence that holds up when a regulator, investor, or acquirer looks closely. That means getting the intended purpose position right before launch, designing a claims architecture that works across clinical, professional, and consumer channels, and sequencing the evidence so it supports both the TGA pathway and a future US filing without starting again. If you are building in the neurotechnology or digital health space and want a clear view of where your product sits, and what it would take to defend it, get in touch.