SKT to Assign Employee IDs to AI Agents and Manage Their Permissions and Responsibilities

[Financial News] SK Telecom said it would establish a system to assign identification information to AI agents used for work and manage the person responsible for each agent, its scope of work and data access permissions. The greater an agent’s autonomy and impact on work, the more stringent the human review and approval procedures will be; performance will be evaluated based on operating records to determine whether the agent should be retained or improved.  
SK Telecom said on the 26th that it had published an “AI Governance Report” covering its AI governance principles, accountability framework, implementation measures and key achievements. The company said it would include not only AI services for users but also in-house AI agents developed by employees or used in their work among the systems it manages.  
According to the report, SK Telecom has designed and is operating an “agent employee ID system” to systematically manage the work performed by workplace AI agents and the people responsible for them. Once an agent has been developed, its basic information is registered, and an employee ID is assigned to agents that meet internal standards in a separate review. Agents assigned an employee ID are tracked by their assigned tasks, responsible parties, affiliations, operating status and performance.
The person responsible for an agent is designated according to the scope of impact and level of autonomy of the work it performs. Users can manage agents that assist with their individual tasks, while agents with greater impact or autonomy are supervised by the person responsible for the work, such as a team leader or executive. The functions an agent can use and the data it can access are also set according to the nature of its work. The greater its autonomy, the more stringent the human review and approval procedures.
During operation, execution status, usage and change history are managed through the “Agent OS” (Agent Operating System). The AI models and tools used by an agent, as well as the process by which it performs tasks, are also recorded so that the cause can be identified if a problem arises. If an anomaly occurs or an immediate response is needed, the agent’s execution can be stopped, and the related actions and change history are recorded.
Agents assigned an employee ID are also evaluated on their operating performance. Their task completion, scope of impact, data access methods and contribution to performance are reviewed to determine whether to retain or improve them, or end their operation. In the long term, SK Telecom is also considering reflecting evaluation results in the organization’s key performance indicators (KPIs).
Customer-facing AI services are assessed for personal information protection, legality, transparency, safety and user control throughout planning, development, launch and operation. Following self-assessments by business units and risk and opportunity assessments by specialist teams, the services are reviewed from the perspectives of personal information protection, legal affairs, security and AI governance. SK Telecom has established a principle of not deploying services that retain risks difficult to accept without mitigation measures.
SK Telecom centrally manages personal information protection and AI governance under the leadership of its Chief Privacy Officer (CPO). The AI Governance Portal tracks risk assessments and review histories for each service, and risks identified when services change or during operation are incorporated into subsequent reviews and improvements.
The mid- to long-term plan calls for building the foundation in 2026, integrating and enhancing the system in 2027, and institutionalizing and expanding it in 2028. In 2026, the company will establish standards for preliminary reviews of AI services and agents, their identification and registration and owner designation, as well as usage permissions and data access scopes. In 2027, it will strengthen differentiated permissions and controls based on autonomy and impact on work, and enhance systems for human review, approval and intervention. The goal for 2028 is to establish a management system covering agents’ operation, evaluation, maintenance, enhancement and decommissioning.
After receiving ISO/IEC 42001 certification in 2024 for its AI management system covering customer-facing AI services, SK Telecom expanded the scope of certification in October this year to include its upgraded AI management system and in-house AI agent management system. Cha Ho-beom, SK Telecom’s CPO, said, “AI governance is becoming even more important in the era of AI agents,” adding, “We will proactively respond to the changing AI landscape and continue to strengthen our AI Trust management.” 
[email protected] Yoon Hong-jip Reporter