- We talked to experts in obesity, tobacco, artificial intelligence, HIV, TB and the NHI to find out what we can expect — and what we can’t — this year.
- The breakthrough six-monthly HIV prevention injection will roll out in April, as we continue to figure out how to deal with the US funding cuts.
- AI is showing great promise, and TB screening is one place it could help in South Africa, where drug resistance is climbing.
South Africa is a study in contrasts. And the 2026 public health agenda is no exception.
Breakthrough HIV prevention medicine has arrived alongside US funding cuts. AI could revolutionise TB screening, but drug resistance is climbing.
New obesity drugs have been launched, but are unaffordable to nearly everyone.
National Health Insurance (NHI) is law, but is facing nine court challenges which may kill it. And tobacco legislation advances – maybe.
We talked to experts about their take on a few of the most pressing public health issues this year. Here’s what they told us.
Obesity guidelines. Check. Anti-obesity drugs. Check. What now?
About one-third of all South Africans are now living with obesity, a condition associated with dozens of noncommunicable diseases (NCDs), such as type 2 diabetes, high blood pressure, kidney disease and some types of cancer.
In November last year, South Africa released its first clinical practice guidelines for the treatment of adult obesity.
It defines obesity as a chronic disease, and helps doctors decide how to treat it, from health support without weight-shaming to medicines and surgery.
It also includes recommendations for GLP-1s (glucagon-like peptide-1 receptor agonists), better known under brand names like Ozempic, Mounjaro and Wegovy.
The medication helps lower blood sugar levels, slows digestion and increases the feeling of fullness. These are approved for diabetes in South Africa, while Wegovy was approved for weight management last year.
Experts say the drugs, which have been used to treat diabetes for 20 years, are a game changer in fighting obesity. But monthly prices that range from about R3 000 to R6 000, according to experts we spoke with, put the drugs out of reach for almost everyone.
They are not yet available in the public sector, and because weight loss is not included under the conditions medical aids are legally required to cover, they have to be paid for out of pocket.
We talked with an endocrinologist who helped craft the obesity guidelines about what’s ahead.
Angela Murphy, endocrinologist
The guidelines will help educate the healthcare sector, like providers, doctors, nursing staff, government and policymakers, about how they approach people living with obesity.
There are many reasons people are obese. But it’s important to understand that up to 70% of the reason a person lives with obesity is genetic. It’s not an issue of willpower. The statistics tell us that a healthy lifestyle alone, on average, can possibly get up to 5% weight loss. This is where medications become incredibly valuable.
The NHI will be stuck in court. And might just die there
After nearly two decades of promises, the National Health Insurance Act was signed into law in May 2024.
The state-run fund promises to do away with the current form of private medical schemes – which about 15% of South Africans have – and better serve the remaining 85% who depend on public health services.
But almost as soon as President Cyril Ramaphosa signed it, lawsuits started piling up. Our latest count: nine separate court actions.
The charges are myriad: procedural flaws in the parliamentary process; inadequate public input; an unrealistic tax-based financial structure; disputes about plans for medical aid schemes; consolidation of control under the health minister; consistent failures in service delivery; and the blocking of access to treatment for asylum seekers and undocumented migrants.
Sitting now with the Constitutional Court is Ramaphosa’s appeal to a previous ruling that required he provide, essentially, a show of his due diligence before signing.
An end-of-year application from the health minister asked to consolidate five of the pending cases and put them on hold until the court’s decision on Ramaphosa’s appeal.
Similar concerns are raised in a South African case study in Global Health Watch, a Lancet-published analysis of global health reform released on January 10.
The experts we spoke with say NHI progress will be stalled – mired in court throughout 2026. That is, they say, if it survives at all.
Susan Cleary, health economist, UCT
What’s been so sad about the NHI debate is that we are saying that an NHI is better than a National Health Service (NHS) system, which is what we’ve got at the moment.
But there are no strong grounds to suggest that NHI is better; these are simply different ways of financing a health system. In fact, the NHS is an outstanding funding model.
Alex van den Heever, economist, Wits University
It’s technically not lawful because it promises outcomes that the system it creates is not capable of delivering. For example, it relies on shifting medical scheme contributions into the tax system at a time when government has no realistic capacity to raise additional tax.
AI: Full speed ahead
Artificial intelligence (AI) is already reshaping public health – diagnosing illnesses, analysing complex datasets and uncovering patterns in research in ways that humans might not even consider.
A study in Nature Genetics found AI was able to predict which people would later be diagnosed with diseases, sometimes years before doctors identified them. AI-assisted screening for TB – which has been tested in South Africa – can read and interpret chest X-rays at a speed and scale humans could never do, with the potential to quickly identify TB hotspots, a cornerstone of prevention.
The experts we spoke with, though enthusiastic and cautious, say AI remains dependent on human expertise and data quality, and they say to researchers and healthcare providers not using it: you better start.
Thomas Hartung, Johns Hopkins Bloomberg school of public health professor and editor of Frontiers in AI
Since the early 2010s, the computing power used to train leading AI systems has been doubling roughly every three months, a pace that can make this year’s AI several times more powerful than last year. Nobody can read the millions of scientific papers that are produced each year, but AI can and does not forget.
Traditional drug development typically takes 12 years and costs roughly $3 billion per drug. AI-first drugs [developed and designed by AI] reached human trials in 2023 after only four years of pre-clinical development. Major companies are now increasingly using AI to speed up every single step of the [development process]; each day saved in getting a drug to market is worth at least $1 million to the developer.
Siphamandla Gumede, researcher, Ezintsha, Wits University
The high-impact goal for 2026 would be to use AI to set off a fully networked public health system that will be able to link the health profile of someone, no matter where they are in the country. So if you went to a clinic and picked up pills, it should be seen up on the system. If you’ve missed your clinic visit, there will be a text, call or email telling you that.
The Tobacco Bill is coming. Just not likely in its current form
Work on tobacco legislation began in 2018. But it took until August 2025 for the Tobacco Products and Electronic Delivery Control Systems Bill to wrap up public hearings. Parliament’s health committee will now decide whether to move ahead.
The legislation bans tobacco product displays at points of sale, single cigarette sales and smoking in public areas – and mostly applies the same rules to e-cigarettes and vapes, which are currently unregulated.
But the industry is pushing back hardest on requirements for plain packaging and graphic warnings (saying it would encourage illicit trade) and the bill’s ban on advertising of electronic devices. The industry uses the term “harm reduction” to argue that the devices are “safer” than traditional cigarettes. But public health experts refute that claim and are worried about vape marketing to children, with use among high schoolers found to be around 17%.
The experts we spoke with are confident that a bill will go through in 2026. But how it ends up is another story.
Lekan Ayo-Yusuf, head, University of Pretoria’s school of public health and director of Africa Centre for Tobacco
We’ve been documenting e-cigarette use since 2010, and the trend is very clear: there is a parallel increase in e-cigarette use and in smoking.
The industry is pushing back on the e-cigarette marketing ban because they say they need to advertise harm reduction to consumers. This is despite studies showing that this didn’t matter to users, especially nicotine-naive youth (people who haven’t used nicotine before) who are increasingly taking up these products for their attractive designs and flavours.
Corné van Walbeek, UCT economics professor and director, Research Unit on the Economics of Excisable Products
Often, people think about illicit trade as counterfeiting or smuggling across borders. But the main source of illicit trade is easy to fix.
SARS could greatly reduce illicit cigarettes with supply control measures, like track and trace technology systems that monitor cigarettes from factory to shop.
TB rates declined. Now we have to keep them dropping
Over the past decade, South Africa has reduced the number of people getting TB by about 60%, driven by better testing and treatment; for every 10 people who used to get TB, only about four do today. Yet we still have one of the highest TB rates in the world. Studies suggest that more than 50% of South Africans carry the infection without being sick, while about one in 10 of those will go on to develop an active disease, usually when something weakens their immune system.
Alongside poverty, poor nutrition and crowded living conditions, the high number of South Africans living with HIV, which leaves them more susceptible to TB, drives our numbers up. Experts are now concerned about the rise in resistance to bedaquiline, the drug that has been used to treat multidrug-resistant TB since 2012. Also worrying: the 15-20% sliced out of our TB programme and uncertainty around research due to funding cuts that could quickly reverse our gains.
But breakthroughs are converging: portable digital X-rays, AI-assisted diagnostics and, since 2022, targeted testing of high-risk groups, like those with HIV, regardless of symptoms. And some research in the pipeline looks promising.
Limakatso Lebina, director of clinical trials, Africa Health Research Institute
South Africa’s first national TB prevalence survey, done in 2018, showed that more than half of people found to have TB were asymptomatic, so not complaining of any symptoms (such as a prolonged cough, fever and unexplained weight loss).
What we don’t know is if asymptomatic people can transmit the TB bacterium. That’s what we hope to find with our research.
Mark Hatherill, director, South African Tuberculosis Vaccine Initiative, UCT
The BCG vaccine, given to children, provides 70-80% protection against TB in children under five. But that protection wears off after about 10 years, which means people can get TB later, as adolescents and adults, who can then infect others. This drives the cycle of the epidemic. To switch it off, we need a vaccine that doesn’t need to be given multiple times.
Right now, there are two leading candidates: M72/AS01E, which is showing 50% protection against the bacterium, and a vaccine called MTBVAC. Both will probably report in 2028-2029.
HIV: Breakthrough science, strained delivery
The public health community is excited about the rollout of lenacapavir (LEN), the groundbreaking six-monthly HIV prevention jab, and recent legal changes for trained pharmacists to prescribe antiretrovirals. But the Trump administration’s funding cuts continue to hamper prevention and treatment.
LEN offers almost foolproof protection for HIV-negative people against the virus, and is expected to roll out by April across 360 government clinics in South Africa. The injections, donated by the Global Fund to Fight Aids, TB and Malaria, will only be enough for 456 000 people – a sliver of what might make a meaningful dent in the estimated 173 000 new yearly HIV infections. Still, it’s an important start.
The health department will need about R2.82 billion to plug the US funding gap for this financial year, but so far the health department has been able to raise only R753 million. The cancellation of 5 800 awards means many clinics were closed – and remain closed – mobile testing units sit unused, more than 24 000 health workers have not been replaced and data systems support is stalled.
Can our stretched system handle it?
Francois Venter, executive director, Ezintsha
Most clinical services for groups of people with a higher chance than the general population to get HIV – sex workers, drug-using populations and LGBTI people – were shut down literally overnight, and have not been replaced. The remaining 25% are in danger from further cuts.
These programmes were successful because US funding was directed at key areas, with money tied to clear actions on the ground. The tiny amount of money mobilised since then has been sent into a vague health department pot with no plan to replace these areas.
Simiso Sokhela, director of clinical research, Ezintsha
These drugs do not work on their own – neither do prevention drugs like LEN nor antiretrovirals. Somebody has to make sure that the drugs get to the people. If the framework is eroded, the drugs will stay on the shelves, or will be given to people who might not take them.
Linda-Gail Bekker, director, Desmond Tutu HIV Centre and principal investigator of PURPOSE 1 trial of LEN
I’m concerned that some people think LEN is a vaccine.
It is important to educate people so they know it is like the daily oral PrEP (pre-exposure prophylaxis) pills or good old-fashioned condoms – but it stays in someone’s body for longer: six months. And after six months, it has to be reinjected, otherwise it does not keep working.
*This story was produced by the Bhekisisa Centre for Health Journalism. Sign up for the newsletter.

 (1).png)

















