About 60% of users say they would be in trouble if generative AI disappeared. Are you relying on AI too much?

Hello there! Thank you for coming to Mugi-ya (´∀｀)

Everyone, I recently saw a statistic that made my heart skip a beat.

According to an additional analysis of a survey conducted in February 2026 and published by ICT Research & Consulting in May 2026, when generative AI service users were asked, "How much trouble would you be in if you could no longer use generative AI?", 18.3% said "I would be in a lot of trouble" and 40.9% said "I would be in some trouble," meaning a total of 59.2% of people answered that they would be in trouble.

However, this survey only measures the feeling of "would it be troublesome if I couldn't use it," not "addiction" in a medical sense. Nevertheless, the fact that about 60% of users feel inconvenienced is proof of how deeply generative AI has integrated into our work and daily lives.

Actually, no one says this, but... the moment I saw this number, I thought, "Oh, that's me."

Mugi can no longer imagine a life without ChatGPT, Gemini, and Claude. Whether it's looking for ideas for articles, brainstorming text, or creating work materials, I realize I'm always with AI. Sometimes I snap back to reality and get a little scared, thinking, "Wait, am I actually doing all of this on my own?" Honestly, there are moments when I feel anxious that this has gone beyond convenience and become a bit of an "addiction."

Don't you all have the same feeling?

I usually work in the business development department of a telecommunications/IT company, handling back-office tasks (accounting, general affairs, legal, and internal controls), and I feel a similar atmosphere within the company. I have a growing sense that more and more colleagues are saying, "I can't create materials without AI" or "It's become natural to ask AI first."

A story about when I relied on AI too much and paid the price

This is embarrassing, but I'll be honest with you...

Previously, I published an AI-generated article for my side job without checking it. The numbers and proper nouns were so smooth, and the writing sounded like it was written by an expert, so I posted it without doubt. But in reality, it contained several pieces of information that were factually incorrect.

The trust from my readers dropped sharply, and for a while, my traffic plummeted. I was so depressed that I seriously considered closing this note... I still clearly remember the feeling of "this might be irreversible" from that time.

Back then, I assumed AI was an "entity that gives answers." But generative AI can output content that is not factual as plausible-sounding text. In technical terms, this is a phenomenon called "hallucination," which simply means "generating incorrect information in a way that looks correct."

The scary part is that the mistake isn't obviously strange, but rather comes out in a way that "seems true." That's why it's hard to notice, and you end up believing it and sharing it. In my case, that caused me to have to rebuild my relationship of trust with my readers from scratch.

Because of this failure, I now thoroughly "check with my own eyes." I learned firsthand that how you interact with AI changes significantly depending on whether you know the risks behind the convenience.

Why have we become so unable to let go?

In the same ICT Research & Consulting survey, there was another interesting piece of data. 41.3% of people answered that their "use of search engines decreased" after using generative AI. In other words, at least for some users, the method of gathering information is starting to change from the traditional search engine-centered approach to one that combines dialogue with generative AI.

Furthermore, the number of generative AI service users in Japan is expected to reach 35.53 million by the end of 2026 and is projected to expand to 51.6 million by the end of 2029. At this point, AI may no longer be just a convenient tool for a few people, but is approaching the status of infrastructure for life and work.

The background to feeling like I'm "relying too much" is that these environmental changes were actually happening. Knowing this number made me feel a little relieved. It meant I wasn't the only person with a weak will (´∀｀)

There is also more concerning data. In a survey by Kore Co., Ltd. targeting 1,008 managers at companies that have introduced generative AI into their operations, 71.3% answered that "work is being hindered by people who cannot use generative AI effectively." Also, in another question asking "What kind of people in the workplace are not using generative AI effectively?", the most common answer was "section managers/leaders in their own department" at 29.3%. In other words, rather than whether or not to rely on AI, "the difference in how one relies on it" may be starting to become a new disparity in the workplace. Note that this survey targeted managers at companies that have already introduced generative AI, and the respondents were skewed toward large companies, so it does not seem to apply directly to all workplaces in Japan. From the perspective of someone in the back office like me, I feel this is not someone else's problem.

Three perspectives to turn dependency from a 'weakness' into a 'weapon'

From here on, I will introduce three ways to handle dependency that I, Mugi, arrived at through my own failures. These are perspectives not often discussed in public, so please use them as a reference.

1. 'Choose' when to use it Instead of leaving everything to AI, decide on a division of roles, such as 'AI for brainstorming, myself for final judgment.' I, Mugi, make sure to always verify primary sources myself, especially for content involving numbers or proper nouns. Conversely, I rely on AI without hesitation when it comes to refining phrasing or generating a large number of ideas.

2. Make fact-checking a 'habit' Treat AI responses as 'reference opinions' and verify important information with multiple sources. In my case, when I include numbers or survey results in an article, I always trace them back to the original publisher to confirm. Since adopting this habit, both the quality of my articles and the trust from my readers have become much more stable than before.

3. Consciously set aside time to think for yourself I feel that if I let AI do too much of the thinking, my own critical thinking skills dull. That is why I make time to write down 'how I would think about it' before asking the AI. Even if it is just for a few minutes, I feel that this extra step allows me to accept the AI's response as my own opinion rather than taking it at face value.

Since I started being conscious of these three things, I feel that I have gradually changed from someone who is swayed by AI to someone who can master it.

Honestly, the number of generative AI users is predicted to continue growing, and I think the percentage of people who feel they would be 'in trouble if it disappeared' will remain high for a while. If that is the case, I think it is much more constructive to be in a state where you can choose 'how to depend on it' rather than fearing the dependency itself.

Within companies, the gap between 'people who can properly master AI' and 'people who have not yet grasped how to utilize it' may become more apparent as a business challenge. That is why I believe that deciding on your own way of handling it, even just one, right now will significantly change who you are in six months or a year.

Summary: Turning anxiety into an opportunity for growth

By incorporating these perspectives, you can expect the following changes:

・Freedom from the anxiety of being swayed by AI ・Gaining trust in disseminating information ・Improvement in research and work efficiency ・Acquiring information literacy that sets you apart from others

I don't think the fact that you can't let go of generative AI is a bad thing at all. It is proof that it has become established in your work and life and is useful. However, high frequency of use and being able to use it properly are not the same thing. What is important is to be in a state where you can explain 'why you are relying on it' and 'how far you judge things yourself' in your own words. If you can do that, dependency will surely become your weapon, not a weakness.

Have you ever felt that you might be relying too much on generative AI? If you have had the same experience, please let me know in the comments. Let's find a better way to handle it together!

If you found this article even a little helpful, I would be happy if you could give it a like or leave a comment!

See you later, Mugiwa!

#TriedWithAI #GenerativeAI #LikeThisPost #Communication #Study #AIDependency #AILiteracy