What Happened
In February 2026, the Therapeutic Goods Administration published updated guidance clarifying how AI-enabled software is regulated as a medical device in Australia. The guidance was covered in detail by InSight+, the online publication of the Medical Journal of Australia, in an article published on 27 September 2026.
The TGA's approach is described as technology agnostic. Whether a product is regulated does not depend on the technology it uses. It depends on what the product is intended to do and what risk that purpose creates.
Phil O'Sullivan, partner and Healthcare Sector Leader at law firm Allens, explained the framework this way: "The question isn't whether something uses AI. The question is: what is its intended purpose, and what level of risk does that purpose create?"
Why It Matters
The distinction has direct consequences for clinicians. O'Sullivan noted that decisions about which AI tools to use "may have implications for patient safety, as well as professional liability, medical indemnity and reputation."
The TGA's regulatory focus falls primarily on the supply of a product rather than on individual clinician users. As O'Sullivan put it: "The regulation is really about regulating the supply of product, not so much an individual clinician user." That said, clinicians who rely on an unregulated or misrepresented tool carry their own professional obligations.
O'Sullivan's position on clinical responsibility is unambiguous: "Your duty of care doesn't pause because you've used an AI tool, so always exercise 100% of your clinical judgment."
Key Details
The TGA framework draws a clear line between tools that make clinical recommendations and those that do not.
AI tools intended to diagnose, prevent, monitor, predict, provide a prognosis for, or treat a condition will generally fall within Australia's medical device framework. Tools that handle administrative tasks sit outside it. O'Sullivan described the boundary plainly: "You can have unregulated products, which are completely fine. Scheduling, billing and document tools, as well as pure transcription or note-taking functionality on its own, are unlikely to be classified as medical devices if they're not making clinical recommendations."
For doctors wanting to check a specific product, the guidance points to a concrete step: "Search the tool on the Australian Register of Therapeutic Goods (ARTG)." From there, O'Sullivan advises clinicians to "see if it's regulated and then compare that to what you're being told about the product in terms of what it's for, what it does, what the purpose is, and how it's being promoted."
He also framed the due-diligence question in practical terms: "Who is telling me that it can do these things? And are they standing behind that?"
Background and Context
The TGA has regulated software as a medical device for some years, but the rapid adoption of AI tools in clinical settings created pressure for clearer guidance. The February 2026 update addressed that gap by spelling out how the existing framework applies to AI-enabled products.
The technology-agnostic principle means the framework does not need to be rewritten each time a new generation of AI arrives. A product's regulatory status is tied to its intended clinical function, so the same rules apply whether the underlying model is a rules-based algorithm or a large language model.
What Comes Next
The InSight+ article notes that understanding whether a tool is regulated can help doctors make more informed decisions about which tools are appropriate to use in their clinical practice. The ARTG remains the primary public reference point for checking a product's status.
No further TGA regulatory changes were announced in the source material at the time of publication.