KT (030200.KS) will partner with artificial intelligence startups to develop new AI transformation (AX) businesses in areas such as digital twins, smart factories and AI kiosks.
KT said on the 20th that it had selected 16 AI startups to take part in its open innovation program, K-PATH 2026, and would begin business cooperation with them.
K-PATH is a program that grooms startups with AI technology as partners for KT's AX business and looks for commercialization opportunities through joint projects. A total of 193 companies applied, and KT made its final selection of 16 based on a comprehensive assessment of technological competitiveness, applicability in industrial settings, commercialization potential and collaborative synergy.
The startups were chosen across seven fields: enterprise AI and agents, physical AI and robotics, data for AI, AI trust and security, AI infrastructure and models, industry-specific vertical AI, and AI ESG. KT said it focused less on raw technological strength than on whether a firm could deliver tangible results by linking up with its existing AX businesses and customer projects.
The project is seen as a strategy to combine the startups' specialized technology with KT's telecom infrastructure and industry-specific customer networks, quickly developing new AX businesses and connecting them to actual customer projects. KT plans to connect the selected companies with relevant business units to verify ontology-based data analysis solutions and to jointly demonstrate AI digital twins in public-sector areas such as disaster response, safety, firefighting and security. It is also reviewing the spread of smart factories using its manufacturing customer network and the joint development of AI kiosks that help older adults with education, ordering and payment. The selected companies include generative AI platform firm Sionic AI, AI agent firm Algorithm Labs, synthetic-data firm SkyIntelligence for robot training, AI security firm Private AI, GPU resource management firm TEN and multilingual AI consultation solution firm CFlat AI.
Separately, KT will begin identifying "quick win" tasks that can produce business results in a short period starting in September. Rather than stopping at technology verification, KT plans to move on to joint proposals to client companies and project execution, then expand tasks with proven results into medium- and long-term joint businesses. Cooperation outcomes and commercialization cases will be unveiled at the K-PATH 2026 event on October 20, and next year the company will pursue a second round of cooperation that includes research and development tasks.
