KT wins 22 financial AX contracts in first half, aims to deliver real results with AI full stack

[Financial News] KT Corporation is speeding up its expansion in financial AI transformation, or AX, after winning 22 AX projects in the financial sector in the first half alone. KT defines the success of AX not as "adoption" but as "results," and says it will help financial firms generate business performance based on its AI Full Stack capabilities spanning data, AI models, platforms and infrastructure.
At the "KT Financial AX Study" held on the 24th at the Four Seasons Hotel in Gwanghwamun, Park Cheol-woo, executive director in charge of financial business at KT Corporation's Enterprise Division, said, "Over the past two years, many companies pushed ahead with AI projects, but many of them stopped at proof of concept, and only 25% moved on to actual operations." He added, "We are observing what projects they are pursuing and why they succeeded."
KT is promoting financial AX as an end-to-end business covering strategy, implementation, operations and advancement. To that end, it has built a "5-layer AX full stack" that jointly develops AX strategies with clients and redesigns work processes based on AI and data. It has also adopted a co-development model with frontline deployment engineers, or FDEs, who work directly on site to solve problems with clients.
Park said, "By the end of the first half, we had won about 22 AX projects in Korea's financial sector alone, a 250% increase from the same period last year." He added, "The main areas are AI Contact Center (AICC), AI agents and data platform construction." KT is currently handling the deployment and operation of AICC systems, including AI chatbots and consultation bots, for four of the country's five major commercial banks. It is also building an AI-based development and operations system, or AIDLC, for some banks. In March, it introduced an agent that supports asset management consultations for private banking staff, and in June it unveiled an agent that customers can use directly for consultations.
In the insurance sector, KT is applying a generative AI-based sales assistant to standardize the consultation know-how of veteran planners so that AI can use it, while also supporting product recommendations and consultations. For credit card companies, it has built an AI data platform that is easier for Large Language Models (LLMs) to use, as well as "AI LLM operations" for stable AI deployment, through an integrated data platform. The project was won in February, and development is under way with a target of launching the service in the second half. Park said, "Finance is an industry where data is well organized and AI's business performance can be clearly measured." He added, "For AX, we are using not only KT's internal LLM but also global frontier models from Microsoft, AWS, Google and Anthropic."
Kim Young-min, head of KT's AX Engineering Division, who followed with a presentation, cited the lack of data, governance and operating systems as reasons why AI projects fail to move beyond proof of concept and into real operations. He explained that in the financial sector, data is scattered across multiple systems, and a large share of the information is unstructured, including documents, terms and conditions, regulations and consultation records, making it difficult for AI to use immediately.
KT therefore sees "data" as a core area within its 5-layer AX full stack. Through its integrated data service, "Data for AI," and its knowledge model, "K-Ontology," KT has converted data into a form that AI can use and established data governance and security systems. Kim said, "In an insurance company project, we structured hundreds of thousands of policy terms and the know-how of top planners into a form that AI can understand." He added, "In financial company projects, we first designed governance on how to manage the data, rather than focusing only on data quality, to reduce repetitive work and employee fatigue."
He also added that the key to AX is deciding what tasks to assign to AI, what data access rights to grant, who will approve AI outputs, how performance will be measured, and how AI will be operated and audited.
Since taking office in March, CEO Park Yoon-young has defined KT as an "AX Platform Company" and made AX business a core growth pillar. More than 800 employees in KT's business-to-business organization are currently working as AX specialists. The company is expanding its enterprise AX ecosystem through partnerships with a range of players, including global consulting firms and law firms.
[email protected] Choi Hye-rim Reporter