Korea's spy agency to release security guidelines on ChatGPT

gettyimagesbank

By Jung Min-ho

The National Intelligence Service (NIS), Korea's spy agency, is working on creating security rules for how to use ChatGPT and other artificial intelligence (AI) chatbots amid growing concerns over the new technology.

According to the NIS on Sunday, it plans to release its guidelines to government departments and agencies later this month. It said it started the project in April in cooperation with experts in the field to take advantage of AI technology while minimizing possible complications.

The leakage of personal or work-related information as well as the creation of realistic “fake news” are among the biggest concerns about the technology, which has proven to be useful in many areas despite some flaws and limitations. With AI reshaping the tech landscape, there has been a mix of expectations and worries about how much impact it will have on people in daily life.

“Developing security regulations is as important as developing the technology itself,” Kwon Tae-kyoung, a Yonsei University scholar who participated in the project, said in a statement. “AI technology should be used on the basis of security policies.”

The intelligence agency said the upcoming guidelines will continue to be updated to protect the national interest against the capabilities AI technology offers to everyone ― including those with ill intentions.

So far, several countries have banned the use of ChatGPT, including Russia, China and Cuba. Italy became the first Western country to join them in March over privacy concerns. But it later lifted the ban after chatbot operator OpenAI implemented changes to satisfy Italian regulators.