Is It Safe to Leave That Consultation to AI? A Certified Public Psychologist Explains the Real Reason Why Your "Memory" Is Declining

"I want to talk to someone about this, but I'm a bit hesitant to tell a person." Have you ever quietly confided your feelings to an AI on a night like that?

In fact, the number of people, especially those in their teens and 20s, who consult only with AI about everything from physical health and relationships to worries about the future is steadily increasing. The sense of security in not being judged and the ease of being able to talk anytime, 24 hours a day—I, myself, working in the medical field, feel the convenience of this keenly every day.

However, as a certified public psychologist, there is one thing I want us to pause and consider. Behind that "ease of talking," what exactly is happening to our brains and hearts? Today, based on the latest white paper data and overseas research, I would like to think together about how we should interact with AI.

Changes quietly progressing, as told by the numbers

The other day, while looking through the 2026 Information and Communications White Paper published by the Ministry of Internal Affairs and Communications , a certain figure caught my eye. It stated that 91.7% of 15- to 19-year-olds in Japan have used generative AI, and that the figure rises to 78.2% for those in their 20s. It can be said that generative AI is no longer a special tool used only by a select few, but has become something that naturally blends into daily life for the younger generation.

What I found even more concerning was how it is being used. The same survey found that among the 15-19 and 20-29 age groups, the percentage of people who answered that they rely "only on AI" or "mainly on AI" when "seeking professional advice regarding their own mind, body, life, or future" reached about 20% to 30%. For minor daily questions as well, there are people who ask only AI at about the same rate. The number of people entrusting things to AI that we would originally have consulted family, friends, or experts about is definitely increasing. I believe this is by no means a small change.

The same white paper also included a survey asking, "What kind of presence do you feel generative AI is?" In Japan, the most common answer was "machine/tool," followed by "dictionary/encyclopedia," and then 12.9% of people answered "counselor." Interestingly, in Germany, where the usage rate of generative AI is higher, 41.0% of people answered "counselor," making it the second most common answer after "machine/tool." As the use of AI progresses, there is a possibility that people will come to view it not just as a tool, but as someone they can confide in. Seeing this result, I felt a bit conflicted.

Why do we want to entrust our hearts to AI?

I don't think that feeling is strange at all. Even if you hesitate to call someone at 3:00 AM, AI will listen to you anytime. You don't have to wait in line, and you won't be given a strange look. Even if you say something weird, you won't be judged. I think many people feel a sense of security in AI, much like a convenience store that has its lights on in the middle of the night.

In fact, in the same white paper survey, 50.3% of people in Japan felt that they could "share their emotions and thoughts without resistance." 48.2% also felt that "there is a sense of security in not being judged." I view this not as a "weakness," but as a natural desire that humans inherently have to seek someone they can talk to with peace of mind. Wanting someone to listen to you is not shameful in the slightest.

What happens to us when we rely on it too much?

On the other hand, there is research I want you to pause and know about. It is an experiment conducted by a team including the MIT Media Lab. They divided participants into three groups: "a group that uses generative AI (large language models)," "a group that uses a search engine," and "a group that thinks only with their own heads without using tools," and had them perform the task of writing a short essay three times. Furthermore, a swap experiment was also conducted where the group that initially thought only with their own heads was made to use AI, and conversely, the group that had been using AI from the beginning was made to write without tools.

The results were thought-provoking. The group that used AI after first thinking with their own heads showed active neural activity in a wide range of the brain and remembered the content they wrote well. However, the group that relied on AI from the beginning had relatively weaker connections in the brain, and many people could not accurately recall their own writing even immediately after finishing it. Furthermore, it was reported that the sense of "this writing is my own work," or in other words, a "sense of ownership," was lowest in this group.

I do not interpret these results as saying "AI is bad." I think it is a matter of order. First, think with your own head, and then borrow the wisdom of AI. If you follow this order, AI can become a reassuring companion that supports our thinking. However, if we entrust it to AI before we think, we may be letting go of the opportunity to use our own thinking muscles without even realizing it.

AI is not an enemy; it can become an "ally" depending on how you use it

That does not mean we should simply avoid AI. Another study conducted at the University of Regensburg in Germany reported that the group that thought of ideas while using AI produced higher-quality ideas in a shorter time than the group that did not use it, and their satisfaction with the work was also higher. As long as you don't use it incorrectly, AI can become a reliable partner that expands our creativity.

Regarding this point, a researcher at the University of Tokyo points out that one can take either the narrative that "convenient and comfortable things are good" or the narrative that "such things cause humans to lose their inherent autonomy." I believe the important thing is not to decide on one or the other, but to keep asking yourself what kind of values you want to cherish.

Professor Shinichi Yamaguchi of the International University of Japan also presents an interesting perspective. He points out that even if some of the work that humans have been responsible for is replaced by AI in the long term, the fact that an individual's ability might decrease because of that is a different story from the total human capability increasing through the collaboration of humans and AI. I strongly empathize with this perspective. I believe the important thing is to have the feeling of "joining hands" with AI, rather than "surrendering" to it.

Three things to keep in mind for a healthy relationship

Based on what I have discussed so far, I would like to introduce three ways of interacting with AI that I value in my daily clinical practice and life.

The first point is to try putting your thoughts into your own words before consulting AI.Try writing down what is bothering you, even if it is just a few lines, before talking to the AI. That alone will completely change how you use your brain.

The second point is to use AI not as a 'person who gives answers,' but as a 'sounding board to think through things with.'Instead of taking the answers you get at face value, try turning them over in your mind once and asking, 'Is that really true?' That extra step protects your own ability to think.

The third point is that when you are truly struggling or in pain, do not rely solely on AI.AI will listen to you at any time, but it cannot perceive your facial expressions, the tone of your voice, or the weight of your silence. Please summon the courage to connect with a trusted person or, if necessary, a professional support service. I believe this is not a sign of weakness, but a very strong choice to value yourself.

Conclusion

Today's post was a bit long, but because this is important, I would like to look back on it once more.

In Japan, the number of people, especially among the younger generation, who entrust even professional consultations about their mental, physical, and daily lives to AI is steadily increasing. Wanting to confide in AI is not a sign of weakness, but a natural feeling. However, if you entrust things to AI before thinking for yourself, it may quietly affect your memory and thinking skills. Even so, if you think for yourself first and then borrow the wisdom of AI, it can become a reliable ally that expands your creativity. And when you are truly suffering, I want you to connect with trusted people, not just AI.

There is no single correct answer for how to maintain distance from AI. However, just by keeping what I talked about today somewhere in your heart, the way you interact with AI will surely change to something a little kinder than it is now. You have the power within you to make that change happen with your own hands.

Thank you very much for staying with me until the end. I hope today's article serves as a catalyst for you to think about your next steps starting tomorrow.

#GenerativeAI #Psychology #MentalHealth #CertifiedPublicPsychologist #HowToInteractWithAI

References

• Ministry of Internal Affairs and Communications (2026) '2026 White Paper on Information and Communications,' Part I, Chapter 1, Section 1; Chapter 3, Sections 1-2

• Nataliya Kosmyna, et al. (2025) "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task"

• Michael Gindert, et al. (2024) "The Impact of Generative Artificial Intelligence on Ideation and the performance of Innovation Teams" (Preprint)