And Sanlam’s 2024 claims data suggests that matters: roughly 24% of all living benefit claims now come from clients under 35; the age group most likely to track their health on a screen long before they ever sit in a consulting room.
However, Dr Marion Morkel, Chief Medical Officer at Sanlam, says the question worth asking isn’t whether these tools are clever. It’s whether they’re being used in the right context.
“AI works best when it complements human judgement, not when it replaces it,” she says. “These tools are fast, convenient and increasingly persuasive. The risk is that we treat the output as the answer, when it’s really just one input among many.”
Why we tend to believe the algorithm
Morkel says one of the biggest risks is how quickly people treat an AI reading as definitive. In clinic proof-of-concept trials, where staff volunteer to test emerging tools such as retinal screening devices or facial analytics for blood pressure, the results can be surprisingly convincing. Even when a clinician repeats the measurement using traditional methods or follows up with additional tests and confirms it was a false positive, people may still struggle to shake off the original result.
That concern becomes more important against the backdrop of South Africa’s evolving health risks. Sanlam Risk and Savings paid out R6.62 billion in individual claims in 2024, up 9.24% from 2023, with cardiovascular conditions making up 20% of death claims and cancer accounting for 54% of all severe illness claims. Prostate cancer claims doubled in men, and breast cancer claims rose 33% in women.
“When we’re seeing this kind of shift, early detection and prevention genuinely save lives,” says Morkel. “But screening is not the same as diagnosis. Many devices marketed as FDA-approved or clinically proven are cleared for wellness screening, not for diagnosing disease. An AI flag is a prompt to follow up with a clinician, not a verdict.”
Why we should track trends, not verdicts
Where AI-powered health tools genuinely shine, Morkel says, is pattern recognition over time, such as sleep, activity, steps, heart-rate variability, and other routine health indicators. With South Africa’s hypertension rate among the highest in the world, wearables that nudge users to know their numbers can be a real ally in heart health.
But she cautions against two opposite dangers. The first is anxiety: AI tools can swing to worst-case interpretations off a single reading or symptom. The second is false reassurance: a comforting wearable result can tempt people to skip routine check-ups or cancel a specialist appointment.
“Both extremes feed into the digital overload picture,” she says. “A constant stream of health metrics can quietly affect mental health and wellbeing, either by ramping up health anxiety, or by lulling people into a sense that they don’t need a real clinician. Neither is accurate, and both can delay the moment someone actually picks up the phone for a proper consultation.”
Guardrails matter
Morkel says professional guidance is increasingly emphasising transparency about how AI is used in clinical decisions, and clearer rules about how personal health data is stored and shared.
“Not all platforms have the same safeguards. Before you share personal information, check what the tool is designed to do, how your data is used, and whether the environment is protected. When AI is part of a health decision, you should be able to ask questions and get a second opinion, not just accept the output.”
The bigger picture for South Africans
Sanlam’s 2024 claims data shows lifestyle-related claims continuing to climb, with stroke and musculoskeletal disorders increasingly affecting women under 50. Additionally, income protection claims among the under-35s have risen from 15% in 2022 to roughly 25% in 2024, a reminder that life-changing illness can strike at any age.
For Morkel, the takeaway is empowerment rather than alarm. “Use AI in context, not in isolation. Let trackers track, not diagnose. And be intentional about trust, transparency and how your data is used.”
Financial protection in a changing world, she adds, is the quieter side of the same conversation. “Health risks are evolving, and the way we prepare for them should evolve too. AI can help you spot a pattern. A clinician can help you interpret it. The right cover can help you carry the cost when life takes an unexpected turn.”