SK Telecom Issues Its First AI Governance Report, With Standards for Internal AI Agents

SK Telecom announced on October 26 that it has published its first “AI Governance Report,” laying out the principles, accountability structures, and implementation plans behind its safe use of AI. The report covers the company’s experience applying those management principles to both customer-facing services and internal operations, along with the main results. SK Telecom said it will build on this work to strengthen its “AI Trust” management strategy, so that customers can use its AI services with confidence.

How AI governance principles translate into day-to-day work

The report focuses on how AI governance principles are actually put into practice. AI governance refers to the framework that determines who is accountable for AI and by what standards it is managed. Alongside its principles, SK Telecom has spelled out a concrete implementation structure. The company said it also presented management guidelines for the new risks that arise as AI agents, which carry out tasks on their own, come into use.

SK Telecom defined its human-centered values for AI in 2021. In March 2024, it followed up by publishing “T.H.E. AI,” its set of AI governance principles. Since then, the company said, it has been fleshing out the structures needed to put those principles into practice across its services and workplaces. This report is intended to gather and share the operating experience and results accumulated over that period.

How customer data is handled

The area that touches customer service most directly is privacy protection. SK Telecom said it runs an integrated management system that combines privacy protection and AI governance into a single framework, overseen by the Chief Privacy Officer (CPO) who leads privacy work across the company. According to the company, whether customer data is used prudently and whether personal privacy is protected are the core review criteria for AI governance.

The company said it has also incorporated “privacy by design” into its work processes, designing privacy protections into a service from the planning stage. The approach involves checking and managing risk across the entire lifecycle of an AI service, from creation through operation.

Internal AI is covered too

For AI agents used in internal operations, SK Telecom said the same management standards apply from development through testing, operation, and decommissioning. In other words, AI that decides on its own how to handle tasks is now within the scope of governance. The report includes management direction and guidelines tailored to how such AI is used. Cha Ho-beom, SK Telecom’s CPO, said that “in the era of AI agents, AI governance is becoming even more important.”

The scope of the company’s international certification has also expanded to cover internal AI management. SK Telecom said that in 2024 it became the first Korean telecom operator to obtain ISO/IEC 42001 certification, the international standard for AI management systems, for its customer-facing AI services. It added that in October 2026, its internal AI agent management framework was certified as well. The company described this as recognition of both the soundness of its AI governance framework and its ability to execute on it.

SK Telecom said it will continue to refine the framework in line with changes in technology and the regulatory environment. The central message of the report is that customer-facing services and AI used for internal work will be governed together under one framework.