Thursday, November 12, 2026

KT to use AI to manage everything from app design to customer communications

Input
2026-11-12 09:25:42
Updated
2026-11-12 09:25:42
KT employees use ‘KDS 2.0’ and the AI-powered ‘Message Generation Agent’ to design digital service screens and review customer notification messages. Photo courtesy of KT.

[Financial News] KT will use artificial intelligence (AI) to automate the design of app and web services and the writing of customer notification messages. The company aims to improve development efficiency and customer convenience by standardizing screen layouts and notification copy that previously differed from one service to another, and applying AI to repetitive tasks.
KT said on the 12th that it had built ‘KDS 2.0,’ a user experience (UX) design system with enhanced AI capabilities, and introduced an AI agent to help draft and review customer notification messages.
KDS sets common standards for screen layouts, design and notification copy used in KT’s digital services. While KDS 1.0, unveiled last year, focused on establishing design principles, version 2.0 has been enhanced so AI can directly read and apply those standards.
To that end, KT built ‘KDS Harness,’ which uses AI to create screens according to design standards and check the results. The system has been applied to My KT, KT.com and Direct Shop, revamped last September, as well as the AI model platform Token Factory.
During the development of Token Factory, planning, design and development were carried out in parallel, while repetitive coding and documentation tasks were automated. As a result, work that was expected to take about eight months using conventional methods was completed in 2.4 months, reducing development time by at least five months.
KT has also introduced AI for drafting customer notification messages. Since KT sends about 170 million text messages each month, it standardized its customer notification copy and introduced the ‘Message Generation Agent’ internally on the 30th of last month. When given key information, the agent generates and reviews copy according to writing guidelines and relevant legal standards, and also supports image recommendations and translation into multiple languages.
According to KT, the time needed to draft and review messages has fallen from 30–60 minutes to about five minutes. Image production time has dropped from four hours to 15 minutes, while translation work that used to take at least two days now takes 30 seconds. In a preliminary test of switching message delivery to RCS, KT’s customer response metrics improved to between 1.3 and 3.4 times their previous levels.
KT plans to unveil a ‘Customer Language Agent’ on the 20th that will simplify difficult expressions used in apps, websites, text messages and other channels. It also plans to expand the use of KDS to external partners and AI devices in the future.
“We will continue to broaden the scope of AX adoption and create a single KT experience that is easy, consistent and trustworthy,” said Manho Won, executive director and head of KT’s AX Platform Division.
[email protected] Choi Hye-rim Reporter