South Korean telecom KT has deployed a suite of AI technologies at Suwon KT Wiz Park, including facial recognition for stadium entry, real-time multilingual translation, and voice synthesis for stadium announcements. The implementation serves as a testbed for KT to scale these AI-driven engagement and operational tools across its broader customer service and media infrastructure.
The successful integration of multilingual AI translation and biometric entry demonstrates how technical infrastructure can directly expand a venue's addressable market and operational security. For the streaming and media ecosystem, KT’s use of prompt-based TTS to maintain brand continuity during talent absences provides a blueprint for scalable, synthetic voice applications in live broadcasting. This shift suggests that future sports rights deals may increasingly hinge on a platform's ability to offer localized, interactive AI layers atop the core feed. Watch for KT to transition these stadium-tested voice and recognition models into its commercial AICC and media content divisions by early 2027.
KT has positioned its stadium AI initiative as part of a broader push to commercialize generative AI across its consumer and enterprise divisions. In early 2025, KT announced a partnership with Microsoft to invest $2.4 billion in AI and cloud infrastructure across South Korea, a deal that includes building Korean-language AI models and data center capacity. That investment signals KT's intent to move beyond connectivity into AI application layers, with the Wiz Park deployment serving as a visible proof of concept for the company's AICC (AI Contact Center) and media platform ambitions. KT's broader AI strategy also includes its Giwa AI platform, which the company has integrated into customer service operations serving millions of subscribers. The Korean sports and entertainment sector has become a testing ground for AI-driven fan engagement beyond KT's efforts. Toss, the South Korean fintech super-app, launched its own AI-powered sports content features in 2025, adding real-time game analytics and personalized highlight generation to its platform. Meanwhile, the Korea Baseball Organization has been exploring league-wide digital transformation initiatives, with clubs investing in smart stadium technologies to attract younger demographics and international audiences. The 57% increase in foreign spectators at KT Wiz Park aligns with a national trend: South Korea's Ministry of Culture, Sports and Tourism reported that KBO League attendance surpassed 10 million fans for the first time in 2024, driven in part by viral social media moments and increased international broadcasting deals. On the technical side, KT's Prompt TTS system represents a specific application of generative voice synthesis that has drawn attention from the broader telecom AI community. KT presented its voice AI research at MWC 2025 in Barcelona, demonstrating multilingual speech synthesis capabilities targeting customer service and broadcast applications. The system's ability to generate natural-sounding stadium announcements from text prompts without requiring pre-recorded audio segments addresses a real operational bottleneck in live venue management. Independent benchmarking of Korean-language TTS systems remains limited, but a 2025 study from KAIST's AI lab found that prompt-based voice cloning models achieved near-human naturalness scores on Korean speech evaluation tasks, suggesting the underlying technology is maturing rapidly for production deployment in live environments.
KT has integrated generative AI, including real-time translation and facial recognition, into Suwon KT Wiz Park. This technical deployment contributed to a 57% year-over-year increase in international attendance. The initiative serves as a proof of concept for KT’s broader strategy to commercialize AI applications in media and customer service by 2027.
The implementation of real-time AI translation services helped drive a 57% year-over-year increase in international spectators at the stadium.
The stadium utilizes facial recognition for entry, real-time translation, generative AI for custom cheer graphics, and a prompt-based Text-to-Speech model for announcements.
KT trained a lifelike Text-to-Speech model using only 10 minutes of voice data from stadium announcer Park Su-mi.
KT plans to expand its facial recognition entry system, developed in partnership with Toss, to all fans by 2027.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source