A recent survey found that 39% of South Koreans have used generative artificial intelligence (AI) tools like ChatGPT. This marks a continuous increase in usage over the past three years, alongside growing concerns about the potential for misinformation and criminal misuse.
The Broadcasting and Media Commission and the Korea Institute for Information and Communications Technology Promotion released the findings of the "2025 Intelligent Information Society User Panel Survey" on May 28.
According to the survey, 38.9% of respondents reported having experience with generative AI, an increase of 14.9 percentage points from the previous year and 26.6 percentage points from 2023. Users spend an average of 49.6 minutes per day using generative AI.
Notably, 68% of respondents reported using generative AI for more than 30 minutes daily.
The primary reasons for using generative AI include its efficiency in information retrieval (86%), assistance in time management (72.6%), support for learning activities (68.2%), help with complex problem-solving (64.8%), and aid in routine tasks (64.8%).
However, concerns about the negative impacts of generative AI have also risen. The survey indicated significant increases in worries about misinformation dissemination (+9.6 percentage points), criminal misuse (+9.0 percentage points), and the generation of content that is difficult to verify (+8.9 percentage points) compared to the previous year. Other concerns, such as job replacement, decreased creativity, copyright infringement, and the creation of biased or discriminatory content, have also grown.
Among those who do not use generative AI, 63.5% cited a lack of knowledge about how to use it or difficulty in learning. Other reasons included the cost of services being burdensome (59.2%), difficulty in selecting suitable services (59.1%), concerns about privacy and personal data breaches (59%), and ethical issues (59%).
The survey emphasized the need for responsible service operations to protect user rights by both generative AI providers and AI recommendation service providers.
For generative AI service providers, 81.3% of respondents indicated that measures should be established to prevent users from generating content that infringes on others' rights, such as internal checks or reporting procedures.
Regarding AI recommendation service providers, 79.3% of respondents stated that users should be informed about the criteria used by AI algorithms to select content. Additionally, about 75% of respondents believed that government intervention is necessary when related services contradict the public interest.
Meanwhile, a survey by WiseApp and Retail reported that as of April, the monthly active users of ChatGPT in South Korea reached 23.45 million, the highest ever recorded. This figure represents an increase of approximately 6 million from the previous year's 17.49 million. Google's Gemini and Anthropic's Claude also saw significant growth, with user numbers increasing by about tenfold and twelvefold, respectively, to 8.45 million and 2.41 million compared to the previous year.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.