Survey on Reskilling Recognition and AI Impact in Japanese Companies 2026 Conducted

Mirai Works, Inc. (Headquarters: Minato-ku, Tokyo; President: Yoshiharu Okamoto) has conducted the 'Survey on Reskilling Recognition and AI Impact in Japanese Companies 2026' and hereby announces the results.

Survey Overview Survey Name: 'Survey on Reskilling Recognition and AI Impact in Japanese Companies 2026' Method: Internet survey Period: March 19-22, 2026 Valid Responses: 400 Target: Company employees (full-time) and executives at companies with 500 or more employees, who have been in corporate planning, general affairs, or HR/labor departments since before 2022 and are involved in personnel training or development.

38.3% of Companies Implement Reskilling Company-Wide; Over 60% in Progress Including Limited Departmental Implementation When asked 'Is your company currently conducting 'reskilling'?', the most common response was 'Implementing as a company-wide measure' at 38.3%. Combined with 'Implementing limited to specific departments/job types' (20.0%) and 'Currently in pilot implementation' (6.3%), it was revealed that 64.6% are engaged in some form of reskilling. On the other hand, 16.0% responded that they 'have not implemented it, including in the past.'

Gap with Government's Definition of 'Reskilling'; Only 9.5% of Initiatives Involve Job Transition When asked 'How does your company perceive the scope of 'reskilling'?', it was found that 61.0% of respondents view it as an initiative that 'does not presuppose a change in duties or roles.' The government defines reskilling as 're-learning to move to growth sectors,' originally referring to company-led initiatives premised on 'labor mobility and job transition.' Based on this definition, the reality is that only 9.5% of companies are practicing 'learning + transition' that aligns with the original meaning of reskilling.

About Half of Acquired Skills are 'Both DX and Non-DX'; Integration of Digital and Business Skills Progresses For those currently implementing, having implemented, or considering reskilling (n=321), when asked 'What skill themes are you trying to have employees acquire (or were acquiring) through reskilling?', the response 'Both DX-related skills and non-DX-related skills' reached 47.0%. This significantly surpassed 'DX-related skills only' (26.5%), suggesting that a refresh of conventional business skills is also required as a foundation to support digital transformation.

Top Priority in DX Education is 'Business Utilization of Generative AI' at 67.8% Among those conducting DX-related reskilling (n=236), when asked 'What are the key themes within DX?', 'Business utilization of generative AI (application, adoption, use case design, business process redesign)' stood out at 67.8%, far exceeding the second-place 'Data analysis/data science' (44.9%). The demand for education on 'mastering' AI is high.

In Non-DX Areas, 'Management' is Top at 33.3%, Followed by 'Corporate Strategy' and 'Marketing' Among those conducting non-DX-related reskilling (n=225), 'Management/Leadership' was the most common at 33.3%. This was followed by 'Corporate/Business Strategy' (30.2%), and 'Marketing' (27.6%).

Reskilling Initiatives Peaked in '2023'; Acceleration Since COVID-19 Pandemic is Clear When asked about the initial launch time of reskilling (n=321), 2023 was the peak at 15.0%, followed by 'before 2019' at 14.0% and 2022 at 12.1%.

Over Half of Companies Forced to Revise Curricula Due to Generative AI's Impact When asked if the emergence and spread of generative AI (since November 2022) affected their company's reskilling (n=321), 37.1% said 'It had an impact, requiring updates to training themes/curricula.' Combined with the 17.4% who said 'It had a major impact, requiring a review of objectives and system design,' over half of the companies were forced to change their measures due to AI's influence.

Biggest Impact of Generative AI is 'Change in Required Skills and Roles' at 48.0% Regarding the specific impact of generative AI's spread (n=252), the top response was 'Business process redesign became necessary assuming AI utilization, changing required skills and roles' at 48.0%. Additionally, 38.9% stated that 'Redefinition of future necessary duties and roles (target profiles) became necessary,' indicating that a fundamental review of organizational rules and roles is underway.

About Half Have 'Already Implemented' Changes Based on Generative AI Regarding changes to measures based on the spread of generative AI (n=321), 'Already changed (partially reflected/in trial)' was the most common at 29.0%. Combined with 'Already reflected as a company-wide standard' (19.9%), about half of the companies have already taken concrete action.

Top Change is 'Review of Skill Themes' (50.0%) When asked what was specifically changed (n=202), 'Skill themes (balance of DX/non-DX and priority areas)' was the top answer at 50.0%. This was followed by 'Target duties/roles' (43.1%) and 'Purpose of reskilling' (38.6%), highlighting that the core parts of education—'what to learn' and 'for what purpose'—are changing with the spread of AI.

Reason for Not Making Changes is 'Lack of Personnel and Budget' at 32.6% For those who have not made or decided on changes (n=95), the main reason was 'Lack of personnel and budget, unable to start changes' at 32.6%. Internal system unpreparedness, such as 'Unable to change because generative AI governance/security/legal arrangements are not keeping up' (28.4%), is also a brake.

'Provision of Learning Opportunities' is the Most Common Policy Element at 58.6%; 40% Introduce 'Skill Visualization' When asked about the elements incorporated into reskilling measures (n=321), 'Provision of learning opportunities such as training and e-learning' was the most common at 58.6%. This was followed by 'Skill visualization is available (skill standards, skill diagnostics, assessments, etc.),' indicating companies' desire to grasp individual skills with data.

Future Policy is 'Medium-Scale Improvement/Enhancement' at 24.6% When asked about the policy for the next 12-24 months (n=321), 'Medium-scale improvement (strengthening business application, KPIs, governance)' was the top response at 24.6%.

Biggest Obstacle to Reskilling Promotion is 'Shortage of Instructors/Mentors' at 25.9% When asked about the biggest obstacle to promoting reskilling (n=321), 'Shortage of instructors/mentors' was the most common at 25.9%. This was followed by 'Personnel/skill data is not organized' (24.3%) and 'Cannot secure time for learning/business application' (21.5%), making 'resource allocation' a pressing issue amid the changing environment of generative AI's spread.

Summary This survey of 400 individuals involved in HR development at companies with 500+ employees revealed that 'reskilling' in Japanese companies is entering a new phase due to the external environmental change of generative AI's proliferation. There is a gap between the spread of the word and the reality; reskilling involving job transition is still in its early stages. Generative AI is no longer a specific specialized skill but is becoming the 'new era's OS (basic literacy)' that all business people should possess, transforming the very roles humans should play. However, the survey also highlighted that internal training systems and data infrastructure are not keeping pace. In the future, a company's ability to face the 'redefinition of duties' based on coexistence with AI, possibly by incorporating support from external professional talent, will be a turning point for its sustainable growth.