{"id":3309,"date":"2026-09-22T08:29:23","date_gmt":"2026-09-21T23:29:23","guid":{"rendered":"https:\/\/news.sktelecom.com\/en\/?p=3309"},"modified":"2026-09-21T16:06:45","modified_gmt":"2026-09-21T07:06:45","slug":"column-from-people-to-people-to-people-to-ai-telecom-remains-at-the-core-of-connectivity","status":"publish","type":"post","link":"https:\/\/news.sktelecom.com\/en\/3309","title":{"rendered":"[Column] From \u201cPeople-to-People\u201d to \u201cPeople-to-AI\u201d: Telecom Remains at the Core of Connectivity"},"content":{"rendered":"
\n

As AI-driven transformation accelerates, the role of telecom operators is evolving and expanding. At the 99th GSMA RCS Group Meeting, held at SKT Tower from September 15 to 18, industry experts actively discussed how AI can be integrated with messaging services and how trust can be established in an increasingly AI-driven communications environment.<\/p>\n

\u00a0<\/p>\n

In this column, Kim Ji-hyung, head of Digital Platform Business at SKT, explores the evolving role of telecom operators in an era when AI agents are becoming increasingly capable of taking action \u2014 not just providing answers. He also examines SKT\u2019s efforts to leverage RCS to advance global standardization.<\/p>\n<\/blockquote>\n

\"\"<\/p>\n

Kim Ji-hyung, head of Digital Platform Business at SKT<\/p>\n

AI is evolving at a rapid pace. It is moving beyond simply asking questions and receiving answers to understanding users\u2019 intent and taking action on their behalf.<\/p>\n

AI agents will increasingly take on tasks that people have traditionally handled themselves, such as making restaurant reservations, comparing products and managing schedules. In the future, a single agent may connect with multiple services and collaborate with other agents as needed to support our daily lives and work.<\/p>\n

This is more than the addition of new features. It represents a fundamental shift in the way people interact with digital services.<\/p>\n

Until now, people have accessed apps and websites directly to find information, verify their identity and select the services they need. Going forward, users will be able to simply tell an agent what they want to accomplish. The agent can then search and compare multiple services and carry out the necessary steps based on the conditions set by the user.<\/p>\n

The Agreements and Rules Needed in an Era of Action-Oriented AI Agents<\/h2>\n

When AI agents simply provide information, the accuracy of their responses is critical. But once they begin taking real-world actions on behalf of people, there are many more factors that need to be considered.<\/p>\n

Who is the agent acting on behalf of? What credentials does it have? How much authority has the user given the agent? If the agent needs to take an action beyond the scope of that authority, how should it ask for confirmation? And how can users verify what the agent has done and the outcome of its actions?<\/p>\n

When people take action themselves, they log in, review the information and provide consent again when necessary. If something goes wrong, they can also explain the steps they took.<\/p>\n

When an agent takes on that role, however, new agreements and rules are needed between people and AI agents. Users need to be able to verify an agent\u2019s identity and credentials, define the scope and level of authority granted to it, and make their own decisions at critical moments. These processes should also be recorded and made available for verification.<\/p>\n

This allows users to remain in control of their choices and authority while delegating tasks to AI agents. The more tasks agents take on, the more important it becomes not only to develop convenient technology, but also to create an environment in which people can use that technology with confidence.<\/p>\n

From \u201cPeople and People\u201d to \u201cPeople and AI\u201d \u2014 Connectivity Remains at the Core<\/h2>\n

Telecom operators can play an important role in this transformation. For decades, telecom operators have connected people with one another and people with services. They have enabled businesses to reach customers through phone numbers, verified users based on subscriber information, and supported access to a wide range of services through messaging and authentication.<\/p>\n

In particular, phone numbers are a means of maintaining a continuous connection with people without being tied to any particular service or platform. As AI agents begin moving across multiple services, the role of phone numbers could take on new significance as a foundation for connecting and verifying agents and their real-world users.<\/p>\n

The role of telecom operators will also extend beyond simply providing connectivity. It could evolve to include verifying who an agent is acting on behalf of, managing the scope of authority granted by users, and ensuring that users and their agents can reconnect when necessary.<\/p>\n

Even when people do not directly access a service themselves, they need to be able to verify that their agent is correctly connected to the services it needs and manage the process. Building on their long-standing experience of connecting people with services, telecom operators are well positioned to support these emerging relationships.<\/p>\n

\"\"<\/p>\n

Building User Confidence Matters More Than Technology Alone<\/h2>\n

As we prepare for the agent era, the key question is not simply how quickly new technologies can be adopted. What matters is how easily users can understand the technology and how confidently they can use it.<\/p>\n

Users should be able to see what an agent has done and what authority it has used. When an important decision needs to be made, users should be able to make the final choice themselves. If a problem occurs, they should be able to review how the task was handled and what the outcome was. They should also be able to modify, suspend or revoke the authority they have granted to an agent when necessary.<\/p>\n

As agents begin carrying out actual transactions, another important question will be how those transactions are conducted and managed. Alongside existing payment methods, digital assets such as stablecoins could potentially be used in emerging transaction environments. Web3 technologies could also be explored as one way of managing an agent\u2019s credentials, authority and transaction history.<\/p>\n

However, simply choosing a particular technology or payment method is not enough. For transactions to be conducted securely, users need to be able to define in advance what requests an agent can act on and within what conditions and limits. When additional confirmation is required, the agent should ask the user for approval again. Once a transaction is completed, users should be able to review the process and its outcome. There should also be mechanisms to place a transaction on hold or modify or revoke an agent\u2019s authority when necessary.<\/p>\n

As the underlying technologies and regulatory frameworks continue to evolve, the priority should not be to treat any single technology as the definitive answer. Instead, it is important to build structures that allow users to remain in control of their choices and authority.<\/p>\n

Such structures and standards cannot be established by any single company alone. As agents connect with one another and interact with a wide range of services, common rules and standards will be needed beyond the technology developed by individual companies.<\/p>\n

SKT is continuing these discussions with a wide range of partners in Korea and around the world. As part of these efforts, SKT hosted the 99th GSMA* RCS** Group Meeting at SKT Tower in Seoul from September 15 to 18.<\/p>\n

*GSMA: The GSM Association
\n**RCS: Rich Communication Services, a next-generation messaging standard that builds on traditional SMS messaging<\/p>\n

The meeting brought together global telecom operators as well as device, platform and messaging companies. During the meeting, SKT hosted a special session titled \u201cAI and the Evolution of Messaging\u201d and proposed a standardization agenda to leverage RCS as a channel for user verification and approval before AI agents execute tasks such as reservations or payments.<\/p>\n

\n