[YONHAP PHOTO = AFP]
Generative AI services including Chat GPT, a chatbot service, use machine-learned data to provide solutions and answers for users. The AI services collect and analyze data inserted by users and collected from the web to improve themselves. Recently, Samsung found out that some generative AI services use data such as exclusive source codes and in-company mail content leaked from their employees.
According to Yonhap News, South Korea's state-funded news agency, on May 2, Samsung sent out messages to its employees in the DX (device experience) Division which combines mobile and consumer electronics businesses that the company will temporarily ban the use of generative AI services on office computers.
"The content entered into the generative AI is transmitted and stored on an external server and is then utilized for AI learning. Therefore, once the content is uploaded, it cannot be retrieved or deleted, posing serious security risks such as the company's confidential information being used as answers to other people's questions," Samsung said in its message.
The tech giant is also found to be in the development of its own generative AI service solution to increase the work efficiency and convenience of its workers. Through an internal survey, Samsung found that about 65 percent of its employees feel that generative AI services had security risks.
reporter
Park Sae-jin
swatchsjp@ajunews.com
