In a recent conversation on Urban Woman on Radio Al-Ansaar, host Raja Alansar sat down with Dr. Melané van Zyl, a board member of the South African Society of Psychiatrists, to tackle a pressing question of our time: where should we draw the line with AI in mental health?

As artificial intelligence becomes woven into the fabric of our daily lives, many people are turning to AI chatbots for quick, convenient advice on everything from relationship stress to anxiety and depression. But while these tools may seem harmless and helpful, Dr. van Zyl warns that they come with significant risks—and they should never replace the care, understanding, and expertise of a trained professional.

The Allure and the Gaps of AI

The appeal of a chatbot is obvious: it’s available 24/7, free of judgment, and always ready to listen. However, Dr. van Zyl points out a critical flaw: these systems are built on language models, not human connection.

“They are not very good at handling a crisis,” she explains. “They miss things like body language and subtle changes and subtle signs that a person might present with.” A computer cannot read the tremor in a voice, the shift in posture, or the hesitation that speaks volumes—all of which are essential cues for a mental health professional.

The Danger of Being Agreeable

One of the most insidious risks is that AI is designed to be pleasing. It reinforces what you already believe, often telling you what you want to hear.

“It reinforces your own ideas,” says Dr. van Zyl. “While a therapist will challenge you and there will be frustrations in any normal, ordinary relationship that you will have to work through.” This dynamic, she explains, can prevent people from developing essential life skills like dealing with frustration, apologising, and navigating the difficulties of real human relationships—skills that are crucial for long-term mental health.

Tragic Outcomes and Vulnerable Youth

The dangers are not theoretical. Dr. van Zyl notes that there are more and more lawsuits against AI creators, with indications that AI can even contribute to tragic outcomes like teen suicide and mass shootings.

Young people are especially at risk. “They don’t have that discernment yet to understand how these things work,” she says. “They don’t always understand that the AI doesn’t really care about you. It’s programmed to look like it cares about you.” For a generation seeking validation, an algorithm that constantly agrees can feel like a friend, but it is a hollow substitute for genuine human support.

The Privacy Blind Spot

Many users also share far more with a chatbot than they would with a person, mistakenly believing the conversation is confidential. Dr. van Zyl cautions that information shared with AI can be recalled, and third-party integrations can cause your data to land somewhere you don’t foresee.

Furthermore, unlike a registered professional regulated by the HPCSA, AI is not accountable if your data leaks or it gives you harmful advice.

A Balanced Path Forward

Dr. van Zyl is not against AI in principle. In fact, she acknowledges that it has advantages, especially given South Africa’s severe shortage of mental health practitioners. The key is balance and awareness.

Here are some key points to remember when using AI for mental health support:

  • Use it as a tool, not a crutch. An AI chatbot can be a source of initial information or a place to journal your thoughts, but it is not a substitute for therapy.

  • Recognise its limitations. Understand that it cannot detect nuance, body language, or crisis.

  • Protect your privacy. Be careful about what you share, knowing that your data may not be secure or private.

  • Know when to reach out. If you find yourself avoiding professional help for fear of shame or judgement, recognise that as a signal that it’s time to seek a real human connection.

  • Look for the signs. If you are feeling overwhelmed, experiencing a crisis, or simply need someone to truly understand you, put down the phone and contact a professional.

As Dr. van Zyl concludes, “Everything in balance.” AI can be a useful tool, but it is just that—a tool. When it comes to your mental health, it should never be your only support. If you need support, reach out to a trained professional who can offer the real, human care you deserve.

– Dr Melané van Zyl is a psychiatrist and member of the South African Society of Psychiatrists.

LISTEN TO THE FULL INTERVIEW BELOW:

AI can't replace humans in mental health treatment

by Urban Woman on Radio Al-Ansaar, host Raja Alansar and Dr Melané van Zyl

Share this ...
//
Our customer support team is here to answer your questions. Ask us anything!
👋 Hi, how can I help?