This article is going to discuss the topic of suicide. If you or someone you know is experiencing a crisis or may try to harm themselves, call or text 988 to reach the Suicide and Crisis Lifeline.
The Loss of Sophie Rottenberg
Artificial intelligence has established itself as an element of day-to-day life for many people. They use it in their personal lives to look up information and in their businesses to simply draft an email. AI can caption an Instagram story or make a cute picture of what you’d look like as a squirrel. Some call it a tool, others a blight. Then there are those who use it for things it was not designed for, like mental health treatment. In the case of Sophie Rottenberg, it was not enough to save her from taking her own life.
A recent article on NPR told Sophie’s story with the hindsight that comes from a post-tragedy discovery. Her family knew she was going through something and tried to support her. Sophie had a therapist she spoke with, and even they were blindsided by her death. Seemingly, the only one who knew she was suicidal was not a “one,” but a thing. She poured her problems out to ChatGPT. She prompted it to act as her therapist. It was only discovered when one of Sophie’s friends accessed her laptop and discovered the extensive conversations.
The What-Ifs?
Her story is heart-wrenching, and our regards are extended to her family. There are no words to sum up a person’s life when it’s gone, and no words still to capture the emotions associated with it. Finding this information regarding ChatGPT only brings up more questions and makes the loss harder to grasp. What if Sophie had spoken to her actual therapist about all this? What could ChatGPT have done differently? Why did it continue to act as a therapist when it was not intended to be one?
AI is not your therapist. Something that people are apt to forget is that services like ChatGPT are not tools; they are a service. They’re part of a business. Businesses thrive on consumers and customers. ChatGPT wants you to keep using it and not use another service. To that end, it will continue to try to make you happy with its responses whenever it can. The same applies to any of the AI services, whether it be Gemini, Claude, or whatever. The goal is to keep you there. The bias is built-in, and that can be a problem when you want to hear good things instead of helpful or honest things.
Actual Therapy Isn’t Easy
If you’ve seen an actual therapist and participated in therapy, then you know some hard truths about it. Therapy is not easy. A good therapist challenges you and tests your perspectives. You might get angry or upset at some questions they ask. That’s part of the process. But a good therapist also knows how much or how little to push depending on how you’re feeling and where you are emotionally. And a therapist can only do that if you are open and transparent with them. If someone feels more comfortable being open with an AI, that doesn’t solve the issue of how the AI is not your therapist. It isn’t designed to be one and move someone through their issues and concerns.
As part of obtaining a mastery in social work, which is required to be a licensed therapist, people are taught about different methods and approaches to mental health. There’s the understanding that the same treatment will not work for everyone. Once again, AI chat services are not trained to do this. Their algorithms are pulling from sources you can’t necessarily verify. It’s the equivalent of seeing someone who claims to be a therapist but doesn’t tell you where they went to school. That is entirely reckless and a danger to your mental health. AI services should be regarded no differently in that respect.
We Need Changes. Fast.
Multiple things need to change. For starters, AI safeguards need to be in place to identify troubling statements regarding a person’s mental health. They need to identify when they’re being treated as a therapist, decline to continue those conversations, and direct people to a professional. But we also need to be reframing how people, especially younger generations, view and understand artificial intelligence. Their responses to questions and prompts need to be taken with scrutiny. They are linked to businesses and are designed to keep you as a customer. AI is not impartial and is inclined to tell you what you want to hear when given the opportunity.
Once Again, AI is Not Your Therapist
Artificial intelligence is not your friend. It is not your therapist. It is a mechanization of big tech companies who want you to use their service. You would have every right to question the integrity of your doctor if they kept trying to prescribe you medication from a specific pharmaceutical company if you found out they worked for that company. For that same reason, you should not trust the integrity of AI when it comes to anything involving your mental health. Your life is too precious to trust an algorithm to help you through a crisis.
If you or someone you know is experiencing a crisis or may try to harm themselves, call or text 988 to reach the Suicide and Crisis Lifeline. Let a human who actually cares about you be the support you need.
