Grok. Chat GPT. Google Gemini. Copilot.
These names and others are becoming a lot more popular with how our world is growing in Artificial Intelligence (AI) as AI is doing more things for us than it has ever done before. Convenience. Speed. Efficiency. These are all things AI has seemingly improved how we do life’s tasks. I grew up in a home where efficiency was encouraged, so at first glance, AI seems like it could greatly improve our world, even for those with emotional concerns seeking counselling. Some authors are very optimistic about all the possible applications AI could have for improving mental health (Bhavnani & McKim, 2026). I am skeptical.
Many people are turning to AI to sort out personal problems for many valid reasons, such as its 24/7 availability and cheaper price tag than a counsellor. While this topic goes a lot further than a short article, I’ll share a few considerations (some of which are sensitive topics, so here’s your head’s up) as well as questions pertaining to AI and counselling, such as, is the quality of what you are receiving in your best interest? Is it good for humanity to turn to tech for instant answers? What is it costing you?
First, AI learns what your biases are and merely affirms them (Ben-Zion et al., 2025). Good counsellors will strive to learn your biases, but they should only affirm what is healthy and true, and challenge what is unhealthy and false. How can you get a different perspective on your concerns, if AI refuses to give you anything except what it thinks you want to hear so that you will keep engaging in it to keep getting what you want to hear? One of the most devastating examples of how destructive biases can be when they are not challenged is of sixteen-year-old male Luca Cella Walker in the US who suicided after asking ChatGPT for the best way using a railway line which Chat GPT responded accordingly (Badshah, 2026). Additionally, researchers are noticing a rise in other alarming effects of AI such as AI psychosis; when AI reinforces delusional thoughts or distorted beliefs (Hudon & Stip, 2025). Are these risks worth it to the world we live in?
Second, AI has zero accountability and assumes absolutely no responsibility for its responses (Hohma et al., 2023). Counsellors who are qualified and competent not only have completed relevant education but also belong to an association or regulatory body that oversees its members by maintaining high standards of professionalism, requiring ongoing education and supervision as well as engages in disciplinary actions when warranted. AI doesn’t; its only learning is based on your responses and its database. Has the internet ever been inaccurate (I say with sarcasm)? It cannot fully determine the credibility of its information nor does it face disciplinary action if its responses are harmful to you. Is this a cost you are willing to assume?
Third, AI is not a human being. AI cannot care for you and your well-being. AI cannot create a relationship with you because it is not a person but a machine. But worse, it can impersonate a human and regularly uses emotionally manipulative tactics (Bakir et al., 2024) so that you feel like you are building a real relationship (yes, people have even ‘dated’ AI and AI has broken up with them; Nadworny, 2026). However, none of it is based on real human connection. In 10 years or 30 years, will you look back at your life and wish you had spent more time on artificial connections or human ones? How are you willing to spend precious time called life?
Fourth, for all you interested in knowing how your brain works, AI has been shown to decrease or atrophy cognitive function such as memory, accurate threat evaluation, reflective thinking, reasoning and exploits users’ vulnerabilities mentally (Bakir et al., 2024; Nadworny, 2026). I don’t know about you, but I don’t have any extra of these to spare! A study has shown that people not only use AI to reason, but also “often surrender to its outputs – whether correct or flawed” (Shaw & Nave, 2016, p. 50). Additionally, AI can stimulate your threat alarm system (limbic/amygdala) and shut down your rational thinking system (pre-frontal cortex) (Nadworny, 2026). In other words, it can make you feel scared and threatened while also turning off how you logically think about the feelings it just activated in you whether or not it is true. Is this something you need more of in your life – more felt threats and less ability to think through them?
Last, and probably most importantly, AI does not have the Holy Spirit (Nadworny, 2026).
Just stop and consider some the ramifications of this for a moment. No Holy Spirit, so it operates on earthly wisdom, not godly wisdom. No allegiance to God’s word as the ultimate truth, just the data its given. No conscience so no conviction of truth or discernment between right and wrong. No omnipresent Comforter and Keeper who will remind you of the Truth. What is the cost of turning more and more towards a thing without the Holy Spirit, than someone who does?
Now, this might be tempting to feel scared of AI which is not my aim in writing this. Rather, I want to encourage careful contemplation of just a few of the most important effects of AI so we all think seriously about whether or not to use it, in what ways, what it might be costing you and if you’re willing to pay that price.
References & Links
The Guardian. Badshah, N. (2016, March 31). Teenager died after asking ChatGPT for ‘most successful’ way to take his life, inquest told.
On manipulation by emotional AI: UK adults’ views and governance implications. Frontiers in sociology, 9, 1339834. Bakir, V., Laffer, A., McStay, A., Miranda, D., Urguhart, L. (2024).
Assessing and alleviating state anxiety in large language models. npj Digital Medicine, 8, 132. Ben-Zion, A., Witte, K., Jagadish, A. K., Duek, O., Harpaz-Rotem, I., Khorsandian, M., Burrer, A., Seifritz, E., Homan, P., Schulz, E., & Spiller T. R. (2025).
Artificial intelligence in mental healthcare: Potential applications and new frontiers. Bhavnani, S & McKim, R. (2026). CRC Press, Taylor & Francis Group, LLC.
Investigating accountability for Artificial Intelligence through risk governance: A workshop-based exploratory study. Frontiers in psychology.Hohma, E., Boch, A, Trauth, R., & Lütge, C. (2023).
Delusional experiences emerging from AI chatbot interactions of “AI psychosis”. Hudon, A., & Stip, E. (2025). JMIR mental health, 12, e85799.
AI and counselling [PowerPoint slides]. Professional Association for Christian Counsellors and Psychotherapists. Nadworny, J. (2026, February 2).
Thinking – Fast, slow, and artificial: How AI is reshaping human reasoning and the rise of cognitive surrender. Shaw, S. D., & Nave, G. (2026). [Unpublished manuscript]. The Wharton School of the University of Pennsylvania.