Half of Japanese Employees 'AI Inexperienced': Widening Skill Gap and Dangerous Reality of Neglected 'Shadow AI'
A survey by SHIFT AI reveals that over half (56.3%) of Japanese company employees do not use generative AI in their work. Challenges include the risk of 'Shadow AI' due to undeveloped internal rules and a widening skill gap. While some experience efficiency gains, security concerns and 'AI fatigue' are also evident.
📋 Article Processing Timeline
- 📰 Published: April 30, 2026 at 22:00
- 🔍 Collected: April 30, 2026 at 13:31
- 🤖 AI Analyzed: April 30, 2026 at 14:16 (44 min after Collected)
Our company, SHIFT AI Co., Ltd. (Headquarters: Shibuya-ku, Tokyo, Representative Director: Shodai Kiuchi; hereinafter: the Company), which operates "SHIFT AI," the No. 1 generative AI learning community with the mission of "Making Japan an AI advanced nation," conducted its second survey on the actual status of generative AI utilization in business among 706 company employees (aged 25-69) nationwide.
Summary of Survey Results
In this survey (n=706), 56.3% of employees responded that they "do not use" generative AI in their work, revealing the reality that "one in two employees is inexperienced with AI."
While expectations for advanced utilization, such as "automation of business processes" and "data analysis," are rising among active users, looking at organizations as a whole, delays in rule development and skill gaps are significant obstacles.
Click here for the previous release >> https://prtimes.jp/main/html/rd/p/000000131.000116644.html
Key Points of Survey Results
1. One in two company employees does not utilize generative AI in their work.
- 56.3% of company employees responded that they "do not use" generative AI in their work.
2. "Shadow AI" risk becomes apparent due to underdeveloped internal rules.
- "Don't know" about internal guidelines: 39.8% / "Clear rules exist": 11.8% / "Experienced Shadow AI": 9.5%
3. Security concerns are the biggest barrier to expanded utilization.
- "Security concerns": 31.7%
4. Polarization is progressing within organizations due to expanding skill gaps.
- "Skill gap": 30.2% / "Advanced users": 7.6% / "Inexperienced users": 56.3%
5. While some feel increased efficiency, it has not led to overall optimization.
- 27.3% of users feel over 1 hour of work reduction per week.
6. Issues in the operational phase, such as "AI fatigue" and differing perspectives, are becoming apparent.
- "AI fatigue": 13% / "Differing perspectives with superiors/colleagues": 12.3%
Detailed Survey Results
How much do you use generative AI in your company's operations?
In this survey, "daily use" increased to 14.7% from 7.6% previously, confirming an expansion of daily active users. On the other hand, "no interest" remains the highest at 36%, and the total percentage of those who do not use generative AI at all is 56.3%. Although a simple comparison is not possible due to the different sample size from the previous survey (n=300), the trend indicates that the base of generative AI users is gradually expanding.
What challenges within your company led to considering generative AI utilization?
"Want to respond to labor shortages" was the highest at 35.1%, maintaining its top position from the previous survey (38.7%). This indicates a persistent expectation for generative AI utilization as a means to compensate for labor shortages amidst persistently high job opening rates. This was followed by "data analysis takes time" (27.5%) and "want to improve support response" (20.8%).
What do you perceive as hurdles to introducing AI into your operations?
"Security concerns remain" at 31.7% and "individual generative AI skills vary" at 30.2% were almost equally ranked at the top. Both figures have increased compared to the previous survey, indicating that as awareness of AI utilization spreads, concerns about security risks and skill gaps are becoming more pronounced.
What tasks do you want to utilize generative AI for in your work?
"Automation of business processes" was the highest at 33.9%, followed by "data analysis and insight extraction" (24.6%) and "support for productivity improvement" (23.9%), confirming a continued high demand for business efficiency and data utilization.
If you were to receive training on generative AI, what content would be ideal?
"Practical application examples and case studies" was the highest at 33.6%. This was followed by "how to utilize generative AI and implementation process" (30%) and "acquiring operational and improvement methods after implementation" (29.7%), indicating a higher demand for practical and continuous utilization skills rather than abstract conceptual understanding.
Are rules and internal guidelines for generative AI use established in your company?
"Don't know" was the highest at 39.8%, revealing that approximately 40% of employees are unaware of their company's generative AI guideline status. The percentage who answered "clear rules exist and are disseminated" remained at only 11.8%, and even when combined with companies where rules exist but are not sufficiently ingrained in the workplace (15.4%), it only reaches 27.2%. "No rules exist and it's left to individual judgment" also accounted for 23.5%, indicating that establishing AI governance is an urgent task for many companies.
Have you had any of the following experiences while using generative AI in your work?
Among those who use AI, while there are positive experiences such as "feeling that work has become easier with AI" (13.7%), there are also negative experiences such as "AI fatigue."
Summary of Survey Results
In this survey (n=706), 56.3% of employees responded that they "do not use" generative AI in their work, revealing the reality that "one in two employees is inexperienced with AI."
While expectations for advanced utilization, such as "automation of business processes" and "data analysis," are rising among active users, looking at organizations as a whole, delays in rule development and skill gaps are significant obstacles.
Click here for the previous release >> https://prtimes.jp/main/html/rd/p/000000131.000116644.html
Key Points of Survey Results
1. One in two company employees does not utilize generative AI in their work.
- 56.3% of company employees responded that they "do not use" generative AI in their work.
2. "Shadow AI" risk becomes apparent due to underdeveloped internal rules.
- "Don't know" about internal guidelines: 39.8% / "Clear rules exist": 11.8% / "Experienced Shadow AI": 9.5%
3. Security concerns are the biggest barrier to expanded utilization.
- "Security concerns": 31.7%
4. Polarization is progressing within organizations due to expanding skill gaps.
- "Skill gap": 30.2% / "Advanced users": 7.6% / "Inexperienced users": 56.3%
5. While some feel increased efficiency, it has not led to overall optimization.
- 27.3% of users feel over 1 hour of work reduction per week.
6. Issues in the operational phase, such as "AI fatigue" and differing perspectives, are becoming apparent.
- "AI fatigue": 13% / "Differing perspectives with superiors/colleagues": 12.3%
Detailed Survey Results
How much do you use generative AI in your company's operations?
In this survey, "daily use" increased to 14.7% from 7.6% previously, confirming an expansion of daily active users. On the other hand, "no interest" remains the highest at 36%, and the total percentage of those who do not use generative AI at all is 56.3%. Although a simple comparison is not possible due to the different sample size from the previous survey (n=300), the trend indicates that the base of generative AI users is gradually expanding.
What challenges within your company led to considering generative AI utilization?
"Want to respond to labor shortages" was the highest at 35.1%, maintaining its top position from the previous survey (38.7%). This indicates a persistent expectation for generative AI utilization as a means to compensate for labor shortages amidst persistently high job opening rates. This was followed by "data analysis takes time" (27.5%) and "want to improve support response" (20.8%).
What do you perceive as hurdles to introducing AI into your operations?
"Security concerns remain" at 31.7% and "individual generative AI skills vary" at 30.2% were almost equally ranked at the top. Both figures have increased compared to the previous survey, indicating that as awareness of AI utilization spreads, concerns about security risks and skill gaps are becoming more pronounced.
What tasks do you want to utilize generative AI for in your work?
"Automation of business processes" was the highest at 33.9%, followed by "data analysis and insight extraction" (24.6%) and "support for productivity improvement" (23.9%), confirming a continued high demand for business efficiency and data utilization.
If you were to receive training on generative AI, what content would be ideal?
"Practical application examples and case studies" was the highest at 33.6%. This was followed by "how to utilize generative AI and implementation process" (30%) and "acquiring operational and improvement methods after implementation" (29.7%), indicating a higher demand for practical and continuous utilization skills rather than abstract conceptual understanding.
Are rules and internal guidelines for generative AI use established in your company?
"Don't know" was the highest at 39.8%, revealing that approximately 40% of employees are unaware of their company's generative AI guideline status. The percentage who answered "clear rules exist and are disseminated" remained at only 11.8%, and even when combined with companies where rules exist but are not sufficiently ingrained in the workplace (15.4%), it only reaches 27.2%. "No rules exist and it's left to individual judgment" also accounted for 23.5%, indicating that establishing AI governance is an urgent task for many companies.
Have you had any of the following experiences while using generative AI in your work?
Among those who use AI, while there are positive experiences such as "feeling that work has become easier with AI" (13.7%), there are also negative experiences such as "AI fatigue."