LG Uplus (032640.KS) will launch an enterprise artificial intelligence (AI) security solution in the second half of this year. As companies accelerate their AI transformation (AX), the strategy directly targets concerns over confidential information leaks from workplace AI use, aiming to solidify the company's position in the B2B security market.
According to reporting by The Seoul Economic Daily on Monday, LG Uplus first applied its in-house developed "U+ Safe AI" internally, focusing on its development department, early this year. The company has recently been coordinating adoption contracts with multiple firms, accelerating commercialization.
"U+ Safe AI" is a product that combines an AI-dedicated DLP (Data Loss Prevention) solution with productivity features. While conventional DLP only blocks employees from removing confidential documents or source code by attaching them to email, USB drives, or cloud services, this solution detects even the act of inputting text into external large language models (LLMs) such as ChatGPT and Claude, pre-blocking the transmission of sensitive information and personal data.
The "enterprise models" provided by ChatGPT, Claude, and others do have protective measures, such as not using input data for training or controlling access permissions. However, the more conservative a company is, the more reluctant it is to hand over internal data to outside operators, repeatedly stalling at the threshold of AX. "Safe AI" aims to break down the barriers to corporate AX entry by automatically screening what may be sent out and what must be blocked.
"The aim is to resolve the inconvenience of companies that either could not use AI at all or had its use excessively controlled," an official from LG Uplus's development team said. "Considering AI's nature of growing smarter as input data accumulates, we designed it to 'block only what needs blocking.' The goal is to automatically monitor and screen information to guarantee employees' AI use as much as possible and boost productivity."
The solution also includes a feature that tallies and manages AI token usage by department on a dashboard. LG Uplus plans to enhance it further to measure AI utilization rates and productivity metrics by organization. "Quantifying AI utilization performance will be especially useful for companies just beginning AX," the official said, stressing that it is "a tool that simultaneously satisfies security and productivity, which are usually considered to be in conflict."
The first target is the financial sector, including banks and securities firms. The fact that financial authorities have recently been gradually easing network separation regulations is expected to work positively. "Financial firms that had been cautious about AX now see they have more room to maneuver and have begun active review," the official said. "After first entering the financial sector, we plan to enhance our LLM-based security technology to control even unstructured data, and expand into manufacturing over the medium to long term."
LG Uplus's emphasis on the security market is interpreted as a move to secure both differentiated competitiveness and a next-generation revenue source. The company foreshadowed the construction of future-oriented security infrastructure last year, successively unveiling its cloud-based integrated account management solution "Alpha Key" and its cloud security solution "U+SASE."
