SK Telecom Co. SKM recently invested $3 million in Twelve Labs, a cutting-edge video understanding company. The initiative aims to bolster multimodal artificial intelligence (AI) applications across a broad range of industries, particularly security, public safety and customer interaction.
Multimodal AI refers to systems capable of processing and integrating data from multiple sources, such as text, speech, image and video. Unlike unimodal AI, which often focuses on a single data type, multimodal AI processes complex inputs to deliver more accurate outcomes.
Twelve Labs, a San Francisco-based startup, is known for its innovative multimodal video foundation models. Its flagship offerings, Marengo and Pegasus, empower enterprises to harness the full potential of their video archives by enabling tasks like semantic search, content summarization, analysis and question-answering. SKM’s decision to invest in Twelve Labs aligns with its broader goal of expanding its presence in the AI domain. The partnership is likely to foster advancements in multimodal AI technology, combining SK Telecom’s expertise in telecommunications and AI development with Twelve Labs’ cutting-edge video understanding capabilities.
A key aspect of the collaboration is the application of multimodal AI in security and public safety. Traditional surveillance systems require operators to monitor multiple CCTV feeds for extended periods, which is time-consuming and exhaustive. Twelve Labs' multimodal AI model enables rapid identification and summarization of significant events, actions and people from any specific day.
SKM’s AI Initiatives Drive Top-Line Growth
SK Telecom continues to gain momentum with its commitment to advancing AI-driven innovations. In November 2024, SKM successfully concluded its AI Customer Service Support System beta service, representing a key advancement in its customer service technology.
In this initiative, the company partnered with global LLM leaders to create an advanced AI system for telecom. Over a year, they restructured their knowledge system, optimized models and built rich training data with input from customer service experts. Through testing and learning, they refined their Telco LLM and large multimodal model (LMM) to suit telecom needs. Powered by SK Telecom's specialized Telco LLM and LMM, the system has demonstrated strong performance since its launch in mid-October.
In the last reported quarter, SKM’s consolidated revenues rose 2.9% year over year to KRW 4.5321 trillion, fueled by healthy growth in roaming and the enterprise sector. Operating income rose 7.1% to KRW 533.3 billion, driven by increased efficiency from integrating AI across the company and strengthened fundamentals through operational enhancements.
For the past three years, SK Telecom has been laying the foundation for becoming a global AI company. This year, it is speeding up its transition by focusing on three revenue-generating areas — AI data centers, enterprise AI and AI consumer services.