As artificial intelligence agents take on a growing share of everyday tasks, the question of how users verify and approve transactions carried out on their behalf is becoming urgent. SK Telecom has put forward a proposal to address this challenge by positioning the RCS standard as a trust layer between AI agents and the people they serve.
The South Korean operator proposed “AI Agent Approval over RCS” as a formal standards agenda item within the GSMA RCS group, the consultative body where global carriers, device makers, platform companies and business messaging operators collaborate on technical specifications for Rich Communication Services. The proposal was tabled during a face-to-face standards meeting hosted by SK Telecom at its Seoul headquarters from September 15 to 18, 2026. Representatives from 16 companies attended, including AT&T, T-Mobile, Vodafone, Orange, Apple, Google, Samsung Electronics and officials from the GSMA Secretariat, with 33 working-level participants in total.
A Verified Approval Channel Built on the RCS Standard
At the heart of SK Telecom’s proposal is the idea that RCS messaging can serve as a dedicated approval channel whenever an AI agent needs a user’s consent before completing a transaction. Today, if the verification step happens entirely inside an AI service, it can be vulnerable to malfunctions or external manipulation. There is also the practical problem that approval methods differ from one AI agent to another, making it difficult for users to judge whether a given request is legitimate.
Under the proposed RCS standard framework, only pre-registered and verified corporations would be able to display their name and logo within the approval message. This means a user would see clear, authenticated branding before deciding whether to proceed. The message itself could include a product image alongside an approval control, allowing the user to review order details and the payment amount and then authorize execution without leaving the default messaging application. Because the approval request and the user’s response remain stored in the smartphone’s conversation list, there is a built-in record that can be reviewed at any time.
Consider a practical scenario: an AI agent arranges a product purchase on a user’s behalf. Before the payment is finalized, the user receives an RCS message showing the item, the price and the name of the verified seller. The user taps to approve, and the AI agent completes the transaction. Convenience is preserved, yet the user retains direct judgment at the decisive moment.
The Importance of a Common RCS Standard for AI and Telecom
The significance of establishing a shared RCS standard for AI agent approvals extends well beyond a single operator. As AI agents begin performing more actions across different platforms, carriers and geographies, inconsistent verification methods could erode user trust and fragment the experience. A common RCS standard would offer interoperability, ensuring that the same approval mechanism works regardless of the carrier, device or AI service involved.
Yoo Kyung-sang, head of SKT AI CIC, said, “The more AI can do on behalf of users, the more important the touchpoint becomes where users directly check critical information and make the final decision.” He added that SK Telecom will work with global partners to make standards discussions more concrete so messaging can evolve into a trust-based channel connecting AI agents and users.
It is important to note that the proposal is currently under discussion as an agenda item within the GSMA RCS group. It has not been adopted as a finalized global standard, and broader deployment of AI agent transaction verification over RCS remains a future prospect rather than present reality.