LG AI Research is accelerating the use of artificial intelligence in public safety services.
LG AI Research said April 30 that it has completed the first phase of development of an “AI Safety Report System” based on its EXAONE model, in cooperation with the Korea Electronics Technology Institute (KETI), and will launch a pilot service within the year.
The system is designed to analyze more than 39,000 safety reports filed each day and automate the full workflow — from intake to screening and classification, transfer to the responsible department, and sending a reply.
The current Safety Report System uses some keyword-based automatic classification, but accuracy can drop when reports include typos or unclear wording. Reports with photos or videos are handled by staff who review the material and then route it to the relevant agency.
LG AI Research said its vision-language model, EXAONE 4.5, addresses those limits by understanding and reasoning over visual information such as photos and videos to improve screening and classification.
KETI plans to use the refined report text generated by EXAONE to further break down key categories, shorten transfer times for high-urgency cases and reduce administrative workload.
For example, if a user uploads a photo of a clogged storm drain, EXAONE analyzes the image and automatically generates the report text. If the case is classified as high priority — such as a flood risk during the rainy season — it can be sent directly to the response department without intermediate steps, enabling faster action.
The two organizations said that as more safety-report data accumulates, they expect to analyze patterns by time, region, type and frequency, and use the results to help develop policies that respond proactively to emerging safety risks.
“Through cooperation with LG, we will be able to reduce the administrative burden and operating costs of public services and strengthen citizen safety,” KETI President Shin Hee-dong said. “We will create a model case of AI transformation that people can feel.”
Lim Woo-hyung, co-head of LG AI Research, said, “By using EXAONE, we will improve both the speed and quality of safety administration and contribute to improving people’s quality of life.”
Separately, LG AI Research said it is also developing “K-EXAONE” and is participating in the Ministry of Science and ICT’s “Independent AI Foundation Model (Dokpamo)” project. It said K-EXAONE earned the top technology score in the first evaluation and ranked first in contributions within the open-source ecosystem.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.