[DigitalToday reporter Chi-gyu Hwang] LG AI Research said on Wednesday it has completed phase-one research and development for the Ministry of the Interior and Safety’s AI Safety Report service with the Korea Electronics Technology Institute (KETI), based on LG’s AI model EXAONE, and is targeting a pilot service within this year.

The Ministry of the Interior and Safety selected LG AI Research’s EXAONE as the AI model for the AI Safety Report project, part of the government policy task to establish a disaster safety management system to ensure public safety.

The AI Safety Report service aims to have AI analyse more than about 39,000 safety reports a day and automate the full process, from receiving reports to screening and classification, transfer to relevant departments and sending replies.

LG AI Research said a keyword-based automatic classification system is currently applied in part, but accuracy drops when reports contain typos or unclear wording. Staff then open and check photos and videos in each report and classify and forward them to the relevant agencies.

In the AI Safety Report service, LG AI Research’s vision-language model EXAONE 4.5 will serve as the brain to screen and classify safety report content that includes photos and videos.

KETI plans to shorten transfer times for important report types and reduce administrative burdens by selecting detailed categories of important reports based on safety report text generated by EXAONE 4.5.

The two sides said that as safety report data accumulates, it will enable pattern analysis based on data by time, region, type and frequency. They said it could be used in the policy-making process to respond pre-emptively to new types of safety risks.

Shin Hee-dong (신희동), president of KETI, said, "Through cooperation with LG, we expect to reduce administrative burdens and operating costs for public services and contribute to strengthening citizen safety." He added, "We will swiftly and accurately handle safety complaints to create a model case for AI transformation that people can feel."

Lim Woo-hyung (임우형), co-head of research at LG AI Research, said, "LG’s direction is to create practical value through expert AI." He added, "Using EXAONE, we will improve the speed and quality of safety administration directly tied to public safety and contribute to improving quality of life."