Originally published in Croakey , edited by Marie McInerney
Computational and AI-enabled systems are, as in other health fields, increasingly embedded across neurosurgical practice, from diagnosis and surgical planning to intra-operative guidance, education, and research.
The inaugural World Conference of Computational Neurosurgery , held in Sydney earlier this month, sought to face not just the unprecedented opportunity these technologies offer, but the ethical, legal, and societal risks.
It resulted in the Declaration of Sydney on the Ethical Use of Artificial Intelligence in Neurosurgery .
Sydney neurosurgeon Professor Antonio Di Ieva, from Macquarie University, was the organiser and president of the conference and played a key role in the development of the Declaration.
In the article below, he outlines the background to and key features of the Declaration, which, he says, offers a roadmap for other fields of health.
Professor Professor Antonio Di Ieva
Antonio Di Ieva writes:
On 15 February 2026, in Sydney, a multi-disciplinary group of global experts met in person for the first time, gathering neurosurgeons and other physicians, engineers and mathematicians alongside lawyers, ethicists, and philosophers.
Amidst busy sessions exploring the latest technological advancements and treatments of brain and spine diseases at the inaugural World Conference of Computational Neurosurgery , there was also an important discussion – to set in place the rulebook for medicine's newest frontier.
Computational Neurosurgery is a trailblazing neurosurgical translational discipline that's moving rapidly into everyday practice, after decades of clinical and translational research.
The field's first-ever ethical debate has delivered outcomes, with over 100 medical leaders signing the Declaration of Sydney on the Ethical Use of Artificial Intelligence in Neurosurgery . It reflects a shared recognition that AI-driven technology is no longer experimental. It is entering hospital workflows, and the time for a formal framework has arrived.
The statistics underscore why new technology is welcome. Cancer Australia data shows that five-year relative survival for brain cancer in Australia is around 24 per cent, having improved only marginally from 21 per cent in the late 1980s. This contrasts strongly with the improvements in lung cancer , where five-year survival has risen from 9.5 per cent to 27.3 per cent over the same period.
The arrival of computational neurosurgery offers us a proven pathway to improving patient outcomes by enhancing surgeries and therapies, minimising symptoms and side effects, and improving care for people living with neurological diseases.
The comprehensive framework delivered at the conference responds to new risks, including the way critical medical decision-making is evolving.
1st World Conference of Computational Neurosurgery in session. Photo: Prof Antonio Di Ieva
'Seeing' brains differently
For more than a century, surgery has operated on a rigid model, with the expert surgeon as a centralised decision-maker, personally reviewing scans, test results and other evidence with a keen eye, before deciding to operate.
Rigorous training, hard-earned knowledge and personal experience drove the life-and-death decisions around surgeries.
But new 21st century tools – driven by algorithms, neural networks and machine learning – are expanding the limits of human observation. Automated software that can review and assess MRI scans and make pixel-by-pixel assessments of images informed by systemic learning has changed the way neurosurgeons 'see' the brain and its diseases.
These systems can now analyse patterns across vast imaging datasets, examining tens of thousands of patient outcomes to quantify risk and uncertainty, and flag alternative approaches that a traditional neurosurgeon might not consider.
The patient benefits are manifest, eliminating uncertainties that prompt biopsies, reducing diagnostic errors, and engaging multidisciplinary care earlier.
The surgeon's role remains critical, but decision-making is shifting from 'one doctor, one assessment' towards AI-driven collaborations where radiologists, oncologists, neurosurgeons, allied health professionals, machine learners, and the patients determine a shared care pathway.
But the new tools also come with serious risks. Accountability for surgical decisions can never be shifted to software, and the handling of patient data in this newly computerised field is paramount.
Ethical risks
As with all new medical breakthroughs, these game-changing tools and techniques need to be managed responsibly by practitioners and their health systems.
The rise of automated 'black box' tools in other fields — software which generate results without providing a full understanding on the way these results were obtained — has demonstrated the terrible impacts on patients when decision-making and review processes break down.
A strong warning lies in the health insurance industry, where major US insurers are facing class-action lawsuits for the use of AI algorithms to override physicians' advice and reject insurance claims.
One lawsuit alleges that an insurer denied more than 300,000 claims in a two-month period, roughly 1.2 seconds per claim. While an extreme example, the controversy dramatically demonstrates how AI tools always require strong ethical oversight and systemic governance.
Neural data, that is, the wavelengths and electrical information inside our most sensitive, personal organ, pose another ethical consideration for neuroscience. Neural implants and brain-computer interfaces promise to restore hearing, sight and even mobility for people with brain injuries and degenerative diseases, and to reduce or even cancel pain.
But these invasive biotechnologies – driven by AI tools – also raise questions about cybersecurity, privacy and personal autonomy. Protecting a patient's neural data from hackers and bad actors – while delivering transformative care – is an issue that practitioners are preparing to grapple with.
The newly signed Declaration of Sydney squarely addresses the major risks and challenges of a complex technology landscape for surgeons, health professionals and their patients. It establishes shared principles across 15 articles:
- patient-centred care and respect for human dignity
- human oversight and clinical accountability
- robust data rights, privacy and mental integrity protections
- transparency, traceability and explainability
- rigorous validation and post-deployment monitoring
- sustainability, equity and global access
- education that preserves foundational competencies
- transdisciplinary and consumer collaboration
- responsible governance of brain-computer interfaces.
Delegates at the World Conference of Computational Neurosurgery in Sydney. Photo: Prof Antonio Di Ieva
Reflecting the fast pace of change, it is structured as a living document, with a custodians' group committed to periodic review as technology, evidence and community expectations evolve.
Importantly, Article 12 refers to the need for sustainability, given training models require large amount of energy (for example, nuclear power plants are in construction to feed large graphics processing units (GPUs) 'farms' that are placed side by side to handle demanding computational tasks) and large amounts of water to cool the systems down.
Therefore, smarter AI architectures and better cured data are required to impact the ecosystems in a sustainable way.
On a personal level, I also strongly endorse Article 13 of the Declaration of Sydney which notes the advancement of computational simulations and refinement of biomodels and biocomputation should be actively supported and prioritised to reduce and, where scientifically feasible, replace the use of animals in neurosurgical research and training.
The collaborative drafting has involved an international taskforce of leading legal thinkers and philosophers, as well as the medical pioneers driving the new field.
The declaration will provide a global framework for neurosurgeons and other practitioners to follow. The core objective is ensuring human oversight remains at the centre of computer-focused medical practice.
And it's no accident that Sydney is where 192 experts from 25 countries met face-to-face and locked the final details down. Australia is recognised not just for our medical and technical innovation, but also for our thought leadership in ethics and legal frameworks in this brand-new world of AI healthcare.
Roadmap for wider health sector
While focused on computational neurosurgery, the Declaration offers something of a roadmap for the wider use of AI in healthcare settings, reflecting a global push for stronger, professional oversight.
An historical parallel is the 1964 Declaration of Helsinki , which created guidelines for clinical research that strengthened public trust and supported a deep investment in innovation.
A strong, global set of ethics for AI across all medical disciplines is needed to protect the public and ensure the door remains open for deeper investment in research, clinical practice and training for the next generation of health professionals.
The Declaration of Sydney is not the end of an ethical journey, but a new beginning.
Declaration of Sydney Task Force: L-R: Sitting: Allan McCay, Katrina Hutchison, Jeffrey Rosenfeld, Margo Somerville, Rita Matulionyte. Standing: Wendy Rogers, Quinlan Buchlak, Antonio Di Ieva, Erico Suero Molina, Aimee DeGaetano, Amin Tavallaii
Hospitals, professional colleges and university educators must review it and begin to follow it.
Researchers, tech innovators and investors must understand it and incorporate its principles into the next breakthrough software or device.
Critically, our governments, regulators and medical boards across the world must embed and enforce it alongside existing laws and rules.
As technology races ahead, it's pleasing to see the collective effort across the diverse field of neuroscience to embrace responsible AI, informed by history and shaped by evolving governance.
While the Declaration of Sydney is a global milestone for computational neurosurgery, it's also pleasing to see Australia play the role of conference host.
After two decades of scientific innovation, our medical practitioners and researchers are blazing another new trail – with ethical standards that will ensure this new discipline is safe, accountable and trusted worldwide for decades to come.
Supporters of the Declaration can sign and pledge online at: https://www.declarationofsydney.ai/
Professor Antonio Di Ieva is an internationally trained neurosurgeon (MD, PhD, FRACS, IFAANS), Professor of Neurosurgery at Macquarie University, Head of the Computational NeuroSurgery (CNS) Lab, lead author of Computational Neurosurgery (Springer) and many other publications , and the 2026 Eccles Lecturer.
He was president of the 1st World Conference of Computational Neurosurgery , and organiser and president of the World Conference of Computational Neurosurgery and the Declaration of Sydney Task Force and Custodians' Lead.
AI disclosure: AI tools were used to assist with writing this article. The author reviewed, verified and approved all content and sources prior to submission.
