The health revolution is real, but SA will win or lose on governance, not gadgets
Healthcare systems everywhere are under strain, and South Africa is feeling that pressure in unique ways: a heavy burden of communicable and non-communicable disease, deep resource gaps between settings and the daily friction of fragmented data and overstretched teams. New technologies promise relief, but they also force uncomfortable questions about accountability and who benefits first.
Image credit: rawpixel.com on Freepik
Head of the Regenesys School of Health Sciences, Dr Rowen Govender, argues that the biggest risk is mistaking technology adoption for system improvement.
“Health tech doesn’t automatically translate into better healthcare. If leadership, governance and workflows aren’t ready, we simply digitise fragmentation and the patient feels no difference,” says Govender.
For Govender, the next phase of health innovation will be determined less by the sophistication of the tools and more by whether health systems can embed them safely with clear clinical accountability, reliable data stewardship, proper procurement and ongoing measurement.
Against that backdrop, Dr Ntukwu Bengu, health policy adviser, National Planning Commissioner and founder of Alchemy Health Technologies, believes we are witnessing a major revolution in health development.
“Even at an early stage, AI is already producing tangible benefits. Processes move quicker, yes, but that there also clearly lots of completely new and in many cases groundbreaking benefits,” says Bengu.
Acceleration of discovery
The most immediate win, Bengu argues, is the acceleration of discovery.
With data-heavy methods and new AI tools, the timeline from research to drug discovery to clinical trials is compressing.
Govender agrees, but he adds a ground-level qualifier about what counts as progress.
“Speed matters, but only when it reduces real clinical risk in terms of missed diagnoses, delayed referrals, duplication, unsafe prescribing.
“If it doesn’t change decision-making and continuity of care, it’s just faster activity, not better outcomes.”
The deeper promise, however, sits in the idea of precision medicine.
Bengu explains: “We’ve been talking about precision medicine for a very long time, but AI makes it now reality because analytics and prediction can surface patterns that conventional examination may miss.
“Traditional medicine is often built around what works for ‘most people’, and averages can only take you so far. AI, in theory, helps clinicians zone in on what the issues are for an individual patient, not a statistical profile.”
Govender is optimistic about that potential and wary of how quickly it can become a new driver of exclusion.
“Precision medicine can’t become a premium layer available only to the already well-served. In SA, the real opportunity is to use analytics to raise baseline care with earlier detection, stronger triage, better follow-up in primary care, so the average patient actually benefits, not only the exceptional case.”
Policy progress
If that sounds like a leap, it’s worth anchoring it in the choices SA is already making at the policy level.
The National Health Insurance (NHI) Act was assented to on 15 May 2024, setting the direction of travel toward more integrated, strategically purchased healthcare services.
A system aiming for universal coverage cannot function well with siloed records, duplicated tests and uneven access to specialist input.
Technology is fast becoming the infrastructure without which policy cannot deliver.
Govender’s point is that infrastructure is not neutral; it shapes behaviour.
“If we build the wrong digital foundations, we lock in the wrong incentives.
“The goal should be trusted data that improves care, with clear ownership, quality controls, and accountability for how decisions are made.”
Yet the harder problem is what happens when AI and human judgment collide.
Bengu anticipates the need for “a responsible way of arbitration” when recommendations conflict.
The clinician remains accountable, but a practitioner who completely ignores algorithmic input “is not necessarily going to be very effective going forward.”
Disciplined decision-making
The future, then, is neither blind trust in AI nor stubborn resistance to it.
It’s disciplined decision-making; knowing when an algorithm is strong (pattern recognition in images, early warning signals in data) and when clinical nuance, context and patient preference must dominate.
Govender’s concern is that disciplined decision-making can’t be left to personal instinct alone.
“We need agreed protocols for exceptions. When an AI flags something, what happens next?
“When a clinician overrides it, how is that recorded and reviewed? That’s how you protect patients and how you protect clinicians.”
Bengu points out that technology can widen inequality before it narrows it.
Robotic surgery is a stark example and is concentrated in better-resourced environments, and clearly for people with medical aid in many cases.
The argument is not to slow down innovation, but to plan for its diffusion.
To learn fast, then democratise this and make it available to everybody, which is also how costs can become manageable over time.
Govender frames inequality as a design decision, not an unavoidable outcome.
“If innovation only happens in the private sector and we hope it trickles down, we build a two-speed system by default.
“From the start, we should be designing for transfer, training models, financing pathways and platforms that can work in constrained settings, not only in flagship hospitals.”
Medtech
Dr ChatGPT: What OpenAI’s foray into health and wellness means for healthcare marketing
Saul Marquez 14 Jan 2026
Automation anxiety
Automation anxiety is real, and Bengu is pragmatic about where risk is highest.
Disciplines already operating heavily in digital environments are more exposed.
For example, radiology certainly has to be near the top of that list, because image-based work is naturally amenable to analytics.
By contrast, high-touch professions like nursing and physiotherapy are less susceptible to outright replacement, though they will still be reshaped by new workflows and tools.
The limiting factor is leadership capacity in health systems.
SA needs managers and decision-makers who can integrate tools, procure them responsibly and measure whether they improve patient outcomes.
“Digital health is a leadership challenge before it’s a technology challenge.
“It’s about change management, governance and making sure the tool reduces burden instead of creating a second layer of admin.”
SA needs measurable innovation that bends the curve on outcomes without deepening inequality.
Better health at scale is the real revolution.