[2026 Generative AI Tool Usage Survey] Microsoft Copilot Tops Corporate Adoption Rate, But Over 80% of Users Cite Insufficient Practical Level

Recently, many companies are rolling out generative AI tools company-wide, and they are becoming standard infrastructure in business settings. However, a gap between the functions provided by the company-wide standard generative AI tool and the actual business needs on the ground is accelerating the practice of departments using different generative AI tools on their own.

Therefore, SDE Partners Co., Ltd. (located in Minato-ku, Tokyo; Representative Director: Taiki Kawamura) conducted a survey targeting office workers at companies with 300 or more employees who use generative AI tools for work. The survey focused on the market share of major generative AI platforms in domestic corporations, the depth of usage by department and job type, and the actual state of tool differentiation.

[Survey Period] March 18, 2026 (Wed) to March 23, 2026 (Mon)

[Survey Method] Internet survey

[Number of Respondents] 1,014 people

[Survey Target] Monitors who responded that they work at companies with 300+ employees and use generative AI tools for work at the time of survey response

[Survey Source] SDE Partners Co., Ltd. (https://sde-partners.com/)

[Summary of Survey Results]

- Adoption share as company-wide standard tools: Microsoft Copilot (45.3%), ChatGPT (45.0%), Google Gemini (28.3%) - Over 80% of users of company-wide standard generative AI tools feel that the functions and accuracy are not at a practical level - Over 70% of users also use platforms other than the company-wide standard generative AI tool - The co-usage rate is ChatGPT (51.3%), Google Gemini (41.2%), Microsoft Copilot (30.1%), Anthropic Claude (19.4%), revealing a dual structure between the company-wide standard generative AI tool and the generative AI tools actually used in practice.

This release presents the latest survey results on the depth of usage by job type, which cannot be seen from adoption share numbers alone, and the bottlenecks of individual optimization that affect organizational productivity.

[Survey Result 1] Current State of Corporate Adoption Share

~Microsoft Copilot and ChatGPT as the top two, with Google Gemini following~

First, we surveyed the adoption status of company-wide standard generative AI tools and found the current state of the platform competition in the corporate market.

The top share is Microsoft Copilot (45.3%), followed by ChatGPT (45.0%), with these two tools almost tied.

While Microsoft Copilot integrated into business applications such as Microsoft 365 is spreading as standard infrastructure, ChatGPT, which is not dependent on a specific platform, still maintains strong support in the corporate market.

On the other hand, Google Gemini (28.3%) remains at about 30%, mainly among Google Workspace users, and is chasing the top two. Notion AI (8.5%) and Anthropic Claude (7.7%) also hold certain shares.

[Survey Result 2] Depth of Usage and Actual Work Content

~Generative AI tools shifting from personal time-saving tools to organizational business transformation~

What specific tasks are the company-wide standard generative AI tools used for?

As shown above, the top three are: creating and editing emails/documents (50.1%), creating outlines of materials (46.0%), and meeting minutes/summarizing information (41.8%).

This indicates that usage as a general-purpose assistant for routine tasks such as document creation and information organization is widespread.

Now, what about the depth of usage?

We evaluated the depth of usage in the respondent's department on a 5-level scale.

The survey showed that Level 1 (Auxiliary) at 23.1% and Level 2 (Established) at 29.8% total 52.9%, meaning more than half use generative AI tools for personal task time-saving. In many companies, AI seems to be permeating as a tool that individuals use occasionally or as a personal time-saver.

On the other hand, about 40% (44.1%) have reached Level 3 (Standardization) or higher, where prompts are shared and the generative AI tool is integrated into the team's workflow, indicating that organizational business transformation has begun.

Furthermore, while companies achieving Level 4 (Advanced) (13.7%) and Level 5 (Automation) (5.4%) are currently limited, as AI agents move to practical stages, it may only be a matter of time before more companies shift to advanced and automated operations.

Next, how does the depth of usage break down by job type?

We scored the depth of usage on a 5-level scale and calculated averages by job type.

The results show clear differences in usage maturity by job type.

Information Systems/IT (2.7) and Legal (2.6) lead overall

The highest score was in Information Systems/IT (2.7), followed by Legal/Compliance (2.6).

Not only IT departments with deep technical understanding, but also legal departments with high needs for text analysis such as document creation and contract review have deeply integrated generative AI tools into their work processes, indicating a more advanced usage stage than other job types.

Management/Marketing (mid-2.5) ahead

Management Planning / Corporate Strategy (2.5) and Marketing / PR / Planning (2.5) form the next upper group. In job types requiring creative and logical output such as strategic planning and content production, the trend of using AI as a thinking partner seems to be taking hold.

Difficulty in Accounting/Finance (2.0)

The lowest score was in Accounting / Finance / Bookkeeping (2.0).

In job types that require per-yen accuracy and primarily handle numerical data, the characteristics of current general-purpose generative AI (LLM) such as hallucination risk pose hurdles, leading to cautious or limited introduction in practice.

[Survey Result 3] Limitations of Company-Wide Standard Tools and the Wall of Practical Level

~Over 80% of users of company-wide standard generative AI tools feel functional insufficiency~

The most notable result of this survey is the practical evaluation of company-wide standard generative AI tools.

When asked whether they feel that the functions and accuracy are not at a practical level while using the company-wide standard generative AI tool, 24.3% answered 'frequently' and 58.7% 'occasionally', totaling 83.0%.