logo
#

Latest news with #Exaone

Korean police bar use of ChatGPT for sensitive data amid leak concerns
Korean police bar use of ChatGPT for sensitive data amid leak concerns

Straits Times

time18-05-2025

  • Straits Times

Korean police bar use of ChatGPT for sensitive data amid leak concerns

As the use of generative AI continues to expand within the force, concerns have been raised about risks associated with inputting sensitive information. PHOTO: REUTERS The Korean National Police Agency has issued a nationwide notice to police stations, warning officials not to input investigative or work-related information into generative AI tools such as ChatGPT, citing concerns over potential data leaks. A KNPA official confirmed to The Korea Herald on May 16, on condition of anonymity, that the agency had sent out a notice titled 'Precautions when using generative AI tools such as ChatGPT' to all police stations. The written directive 'emphasises that investigative or work-related data as well as personal information of those involved in the case should not be entered when working with generative AI,' the official explained. The agency also advised police officers to 'refrain from responding to questions asked by generative AI that are related to security' and to 'ensure thorough review measures have been taken when implementing AI for investigative projects related to information technology', they added. When asked if investigators often use generative AI, another investigative police official, surnamed Kim, told The Korea Herald that 'using generative AI like ChatGPT has become common among many young police officials in recent years.' 'I know that ChatGPT, for example, has helped many police officials, including myself, when reviewing relevant laws or investigative cases from the past that we may not be aware of,' Kim continued. In March 2023, the KNPA also stated that it would work with ChatGPT to draft essential documents in English when dealing with cross-border crimes to enhance efficiency. At the time, the KNPA stated that there were no concerns of personal data or confidential investigative information from being inputted, as ChatGPT would only be used to generate documents in English. As the use of generative AI continues to expand within the force, concerns have been raised about risks associated with inputting sensitive information, including personal data or confidential matters related to police investigations. A senior police official told JoongAng Ilbo, 'If investigation records from an internal police network are transferred to external AI-generated websites, the risk of personal information leakage or the exposure of confidential investigation details increases.' 'Using ChatGPT to refine parts of investigative reports is like letting important investigative information accumulate beyond the police's control.' In a response to such concerns, the KNPA confirmed that it has been working with LG CNS since the beginning of 2025 to develop a generative AI model catered to aiding police officials in investigations, using LG's AI model, Exaone. Once the investigative, AI-based support service is fully developed, the service will be implemented into the internal system of the police force, according to the KNPA. It will be able to summarize written statements or show similar cases based on the type of crime while analyzing underlying issues of an investigation to help officers pinpoint and focus on important aspects during their investigations. The AI service will also assist in drafting investigative documents. THE KOREA HERALD/ASIA NEWS NETWORK Join ST's Telegram channel and get the latest breaking news delivered to you.

Korean police bar use of ChatGPT for sensitive data amid leak concerns
Korean police bar use of ChatGPT for sensitive data amid leak concerns

Korea Herald

time16-05-2025

  • Korea Herald

Korean police bar use of ChatGPT for sensitive data amid leak concerns

KNPA, LG CNS working to develop generative AI specifically catered to police investigations The Korean National Police Agency has issued a nationwide notice to police stations, warning officials not to input investigative or work-related information into generative AI tools such as ChatGPT, citing concerns over potential data leaks. A KNPA official confirmed to The Korea Herald on Friday, on condition of anonymity, that the agency had sent out a notice titled 'Precautions when using generative AI tools such as ChatGPT' to all police stations. The written directive "emphasizes that investigative or work-related data as well as personal information of those involved in the case should not be entered when working with generative AI," the official explained. The agency also advised police officers to 'refrain from responding to questions asked by generative AI that are related to security' and to 'ensure thorough review measures have been taken when implementing AI for investigative projects related to information technology," they added. When asked if investigators often use generative AI, another investigative police official, surnamed Kim, told The Korea Herald that 'using generative AI like ChatGPT has become common among many young police officials in recent years.' 'I know that ChatGPT, for example, has helped many police officials, including myself, when reviewing relevant laws or investigative cases from the past that we may not be aware of,' Kim continued. In March 2023, the KNPA also stated that it would work with ChatGPT to draft essential documents in English when dealing with cross-border crimes to enhance efficiency. At the time, the KNPA stated that there were no concerns of personal data or confidential investigative information from being inputted, as ChatGPT would only be used to generate documents in English. As the use of generative AI continues to expand within the force, concerns have been raised about risks associated with inputting sensitive information, including personal data or confidential matters related to police investigations. A senior police official told JoongAng Ilbo, 'If investigation records from an internal police network are transferred to external AI-generated websites, the risk of personal information leakage or the exposure of confidential investigation details increases.' 'Using ChatGPT to refine parts of investigative reports is like letting important investigative information accumulate beyond the police's control.' In a response to such concerns, the KNPA confirmed that it has been working with LG CNS since the beginning of 2025 to develop a generative AI model catered to aiding police officials in investigations, using LG's AI model, Exaone. Once the investigative, AI-based support service is fully developed, the service will be implemented into the internal system of the police force, according to the KNPA. It will be able to summarize written statements or show similar cases based on the type of crime while analyzing underlying issues of an investigation to help officers pinpoint and focus on important aspects during their investigations. The AI service will also assist in drafting investigative documents.

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into the world of global news and events? Download our app today from your preferred app store and start exploring.
app-storeplay-store