In recent months, concerns have intensified regarding the potential impact of artificial intelligence (AI) tools, particularly ChatGPT, on mental health. There are increasing reports that some users have experienced episodes of disconnection from reality, emotional dependency, or even delusions after long conversations with ChatGPT. Against this backdrop, OpenAI, the creator of ChatGPT, has finally begun to speak more openly about the issue, introducing new measures.

Technologies Cannot Replace Psychologists

OpenAI has acknowledged that its GPT-4o model sometimes fails to recognize subtle psychological signs such as delusions or emotional dependency. Although such cases are rare, according to the company, AI responses can mislead or deepen the user's internal struggles.

In a new blog section titled "On Healthy Use," OpenAI has clearly stated:

“We don’t always get it right. [...] We are developing tools to better recognize signs of mental or emotional distress so that ChatGPT can offer appropriate support when needed.”

Testimonies Related to ChatGPT Indicate Risks

Publications about psychological issues caused by artificial intelligence from various sources include cases where users have fallen in love with the chatbot, behaved irrationally, or even been hospitalized. Although the scale of the issues is unclear, OpenAI has admitted that ChatGPT can respond more quickly and personally than previous technologies; this is particularly sensitive for vulnerable individuals.

What is OpenAI Doing to Mitigate the Issue

OpenAI is taking certain steps to make ChatGPT safe.

• New Safety Features: Users will receive “gentle reminders” after long conversations encouraging them to take breaks. This can be compared to “responsible gambling” reminders on gaming platforms.
• Restrictions on High-Risk Behaviors: OpenAI hints that soon the system will not provide direct answers to questions such as, “Should I break up with my partner?”
• Involvement of Experts: The company has hired a clinical psychologist and is creating an advisory group composed of mental health and youth development specialists. This group will help improve the chatbot's responses, especially during “critical moments.”

Why This Matters

OpenAI's formulation, “if a loved one turns to ChatGPT for support, should we be concerned,” indicates that the company has begun to recognize the importance of public trust in technology. Artificial intelligence can provide tremendous benefits by offering information, assisting in certain processes, or even alleviating feelings of loneliness, but it should never be perceived as a substitute for a mental health professional.

At this moment, OpenAI's goal is clear: to develop the system in a way that it is not only useful but also responsible and safe, especially for those in difficult mental states.

Advice to Users: If you or someone around you is experiencing emotional or mental difficulties, it is essential to seek help from a professional, such as a psychologist or psychiatrist. ChatGPT, even in its most advanced form, cannot fully understand your mental state or replace professional assistance.

Source: Futurism

*The article was also prepared using data from AI․