Megazone Cloud has released the results of a survey on AI usage among 749 AI and IT personnel from domestic companies. The findings reveal that when companies utilize AI across their organization, concerns about return on investment (ROI) uncertainties are only 13.1%. This is significantly lower compared to 34.9% of companies planning to implement AI within 1-2 years who expressed concerns about ROI uncertainties.
On the 26th, Megazone Cloud, in collaboration with technology media and research company Foundry (formerly IDG), shared these findings. According to the survey, 55.7% of domestic companies are already using generative AI either across the entire organization (22.4%) or within specific departments (33.2%). Including companies currently in the process of implementation or planning to do so within the next 1-2 years, it’s projected that over 85% of companies will use generative AI in their operations by 2026.
Large companies have a company-wide adoption rate of 35.1%, more than twice that of small and medium-sized enterprises. Among industries, IT and telecom/broadcasting sectors show the highest adoption rate at 37.5%.
Additionally, 60.3% of respondents rated generative AI as a high priority in their IT strategy, and 74% reported an increase in related investments compared to the previous year. Of those who considered it a very high priority, 48.1% said investments have increased significantly; 37.6% of companies that considered it a moderate priority reported slight increases in investments.
Despite the rapid adoption of generative AI by many companies, concerns about trust and security remain high. The most significant concern is the generation of incorrect information and lack of result reliability (61.3%), followed by security and privacy risks (53.3%). Other concerns include copyright and legal liability issues (21.0%) and ROI uncertainties (19.8%).
While the adoption of generative AI is spreading quickly, many companies still face challenges regarding execution capabilities, such as securing talent and technology. Nearly half of the surveyed companies (49.8%) cited a lack of technical personnel and expertise as the biggest obstacles to adopting generative AI. Other challenges include difficulties in securing proper infrastructure and data (32.0%) and lack of management support and investment (21.0%).
The most frequently cited necessity for successfully expanding generative AI use in companies was strategic intent and leadership from management (41.1%).
By 2026, 79.3% of companies expect to increase their budgets related to generative AI. Of these, 26.8% anticipate a 20-49% increase, while 38.7% expect a modest increase of less than 20%. Companies that significantly increased their investments this year reported that they plan to expand their budgets by more than 50% in 2026, indicating a strong intention to increase investment among companies where generative AI use is more established.
The main purposes for using generative AI include improving work efficiency and productivity (70.5%). Key departments utilizing it are software development (44.1%), IT operations (40.3%), marketing, customer service, and research and development. Common usage types include document summarization and report writing (43.1%), data analysis and insight extraction (40.3%), and programming assistance (37.0%).
Megazone Cloud’s Chief AI Officer Kong Sung-bae stated, “The survey results show that the biggest challenges to the practical spread of AI are the limitations in skilled workforce and infrastructure, and issues of security and privacy protection.” He emphasized the importance of partnering with external experts who have accumulated experience and technology to effectively overcome these challenges.
He added, “Based on the experience accumulated from collaborating with various clients, Megazone Cloud provides strategies from setting clear goals to achieving ROI when implementing and operating generative AI. We aim to strengthen our role as a customized partner by offering comprehensive support, not just in technology and infrastructure but also in data security and privacy protection.”