Announced details
Work-related use by employees, corporate information management and cybersecurity
67% of respondents (201 people) use generative AI for work, approximately 15% use shadow AI, approximately 60% enter work-related information, approximately 35% have experienced near misses, approximately 14% have experienced an actual problem, approximately 37% often or sometimes use output without checking or correcting it, approximately 43% rarely or intentionally do not share the fact that they use AI, and approximately 45% have no AI usage rules or do not know whether such rules exist
In this announcement the company addresses none of Contract Value / Offering price / Impact on Business Performance.
An internet survey by Cyber Security Cloud,Inc. (survey period: 21–22 April 2026; fieldwork contractor: Rakuten Insight, Inc.; valid responses: 300)
| Item | Details |
|---|---|
| Names of Products / Services Covered | Generative AI (Microsoft Copilot, Google Gemini, personal ChatGPT accounts, and corporate ChatGPT contracts) |
| Applicable area | Work-related use by employees, corporate information management and cybersecurity |
| Quantitative Effects | 67% of respondents (201 people) use generative AI for work, approximately 15% use shadow AI, approximately 60% enter work-related information, approximately 35% have experienced near misses, approximately 14% have experienced an actual problem, approximately 37% often or sometimes use output without checking or correcting it, approximately 43% rarely or intentionally do not share the fact that they use AI, and approximately 45% have no AI usage rules or do not know whether such rules exist |
| Source | An internet survey by Cyber Security Cloud,Inc. (survey period: 21–22 April 2026; fieldwork contractor: Rakuten Insight, Inc.; valid responses: 300) |
Items this document does not state (14)
These are items the company does not state in this document. We do not leave them blank, nor fill them in from other sources or by inference.
The reasons the company gives
The company's own explanation, not something inferred from the numbers.Insufficient verification of generative AI output, overconfidence in generative AI, non-disclosure of AI use, and the absence or lack of clarity of usage rules were cited as contributing factors.