[2024 Mirae Company Forum] "For Successful AI Adoption in Companies, Data Security Must Be Addressed"

'Why AI Transition?' Panel Discussion
Concerns Over Confidentiality Leaks When Using Open Source AI Tools
Security Issues Must Be Resolved for Full AI Adoption
Need to Establish Data Governance Raised

Artificial intelligence (AI) experts emphasized that companies must first address the ‘security’ issue to adopt AI. Although there are generative AI tools like ChatGPT that anyone can use, there is a risk of sensitive corporate secrets being leaked during their utilization.

At the second session of the ‘2024 Asia Future Enterprise Forum’ held on the 22nd at Lotte Hotel in Jung-gu, Seoul, Lee Jin-hwan, Software Campus Head (Executive Director) of automotive parts company HL Mando, mentioned, "Security is one of the biggest concerns in vehicle software (SW) development due to the transition to software-defined vehicles (SDV)." As a Tier 1 parts supplier, HL Mando serves automakers competing in the global market, making the maintenance of security for each client’s orders paramount.

Lee said, "Because we must maintain client security, it is difficult to directly use open-source-based AI in publicly developed software," adding, "We are striving to utilize lightweight language models (sLLM) that process specific data and integrate them into our in-house cloud platform through a customized process."

The second session of the forum was conducted as a panel discussion on the theme ‘Why AI Transformation?’ The moderator was Jang Byung-tak, Director of the AI Research Institute at Seoul National University. Panelists included Kwon Young-jun, Head of Samsung SDS Research Institute (Vice President), Kim Min-su, AI and Big Data Management Leader (Executive Director) at CJ Logistics, Ham Cheol-hee, Master (Executive Director) of Samsung Electronics Visual Display Business Division, and Lee Jin-hwan, Executive Director of HL Mando.

On the 22nd, a panel discussion on the theme "Why AI Transformation?" was held at the '2024 Asia Future Business Forum' hosted by Asia Economy at Lotte Hotel in Jung-gu, Seoul. From the left: Jang Byeongtak, Director of AI Research Institute at Seoul National University; Kwon Youngjun, Head of Samsung SDS Research Institute; Kim Minsu, AI and Big Data Management Leader at CJ Logistics; Ham Cheolhee, Master of Samsung Electronics Visual Display Division; Lee Jinhwan, Head of Software Campus at HL Mando. Photo by Kim Hyunmin kimhyun81@

Samsung SDS highlighted the necessity of ‘data governance.’ Data governance refers to establishing policies and processes to manage vast amounts of data, encompassing data quality and security. Vice President Kwon stated, "Companies must build data governance on how to collect and utilize data and determine which data should be confidential," adding, "Establishing a system for various permissions will enable successful AI training and fine-tuning."

He then identified data, security, cost, and performance as key factors for successful AI transformation in companies. He added, "Many AI models are open source, and methodologies are often publicly available," emphasizing, "Specialized data will be a company’s competitive edge and its greatest asset."

The panelists predicted that AI’s role in future society will evolve beyond simple analysis and prediction to replacing human labor. Professor Jang, known as a top domestic expert in machine learning, explained recent changes in AI’s role. He said, "In the past, AI was mainly used for analysis, classification, and prediction, but recently, generative AI has achieved significant innovation by directly creating content," adding, "Now, active research is underway on action-oriented AI that performs physical tasks and labor."

Kim Min-su, Management Leader at CJ Logistics, mentioned fully autonomous truck transportation in this context. He explained, "As truck drivers age, risks inevitably increase," adding, "The pilot operation of fully autonomous truck services on highways and the use of various robots inside logistics warehouses using AI are in the same vein."

CJ Logistics partnered with autonomous driving startup Mars Auto to start a pilot project in March. An 11-ton large truck operates six times a week on a 218 km trunk route from CJ Logistics’ Incheon Terminal Center to Okcheon Hub Terminal. Inside the logistics warehouse, robots equipped with AI deep learning technology assess box volume and the amount that can be lifted at once to find and execute the optimal work sequence.

Jinhwan Lee, Head of HL Mando Software Campus, is giving a lecture at the '2024 Asia Future Business Forum' hosted by Asia Economy on the 22nd at Lotte Hotel, Jung-gu, Seoul. Photo by Hyunmin Kim kimhyun81@

HL Mando sees the potential for AI utilization in the automotive industry’s transition to SDV (software-defined vehicles). As software usage in automotive parts increases, development cycles accelerate, and challenges in quality assurance grow. HL Mando decided to actively use AI to reduce manpower and development time in the software development process.

Lee said, "In the automotive field, AI has expanded beyond a tool for autonomous driving to a means of software development processes and product manufacturing," adding, "We are selecting areas where AI utilization has the highest value and integrating AI into our in-house platform."

AI application is also expanding in home appliances used in daily life. Ham Cheol-hee, Executive Director at Samsung Electronics, said, "AI cannot be left out when talking about TVs," adding, "We will increase the application of on-device AI in TVs." The biggest change viewers can feel by incorporating AI into TVs is image quality control. AI analyzes low-resolution content and autonomously adjusts brightness and blurring to convert it into high-resolution video.

Ham explained, "AI is used to convert low-quality, low-resolution videos into high-quality, high-resolution videos," adding, "We are especially researching methods to analyze each video and perform optimal upscaling according to 512 categories."