AI in healthcare: Trust is growing, but human oversight remains key

As artificial intelligence becomes more common in healthcare, a key question remains: can AI be trusted with your health?

Experts say public confidence has grown in recent years, but stress that safeguards are still essential.

A study published in the British Medical Journal has raised serious concerns about AI chatbots like ChatGPT, Gemini and Grok, finding that nearly half of their health-related responses are inaccurate or incomplete.

ALSO READ: Africa & Artificial Intelligence: 'For AI to work, people need to understand it'

According to Braden van Breda, CEO of AI Diagnostics, AI is increasingly reliable in areas like medical diagnosis, where it can process large amounts of clinical data and support healthcare workers in identifying conditions.

He said trust in the technology has improved since 2020 as its use in real-world healthcare has expanded.

"I think a lot has changed, especially public opinion on what AI can do. Obviously, there are guardrails that one needs to keep in place to make sure that it's trustworthy in general. I think a lot has changed over six years for all in that regard."

Van Breda added that AI is already helping nurses and clinics, particularly in underserved areas, by improving access to diagnostic support and speeding up referrals.

However, he stressed that AI is best seen as a support tool for healthcare workers, rather than a replacement for human medical judgment.

To listen to van Breda in conversation with 702 and CapeTalk's Africa Melane, use the audio player below: