Datadog Korea Head Eom Suchang: "Visibility Is Key to Security in the AI Era... Governance Must Be Built Through Proactive Detection"

Gaps in AI-Based Cyber Threat Response Systems
Emphasis on "Unified Visibility" Across Applications and Infrastructure
Evolution from Monitoring to Autonomous AI Response

"In the AI era, we need to integrate applications, infrastructure, and data to detect anomalies across entire systems and take action before problems arise."

Eom Suchang, head of Datadog Korea, speaks in an interview with The Asia Business Daily on the 27th. Photo by Lee Eunseo.

Eom Suchang, head of Datadog Korea, emphasized in an interview with The Asia Business Daily on October 27 that governance is needed to support preventive measures and rapid identification of root causes amid a series of hacking incidents involving South Korean companies. As AI diversifies attack methods, it has become harder to block threats using individual security systems and reactive measures alone. He said companies need to establish governance by integrating development, operations, and security data. "As AI adoption accelerates, gaps are emerging as companies work to establish response systems," Eom said.

Datadog is a security observability company that integrates companies' application, infrastructure, and log data to help them assess service health and identify the causes of problems. Established in 2019, Datadog Korea had secured more than 1,000 customers across sectors including finance, manufacturing, commerce, and gaming as of August this year.

AI is broadening the range of threats companies need to prepare for. They must consider not only attacks that disrupt services or steal data, but also attempts to contaminate data used by AI and manipulate agents into making incorrect decisions. "Companies need systems to verify what data AI is using, whether agents are making decisions within their authorized scope, and whether they can filter out incorrect information," Eom said.

The approach he advocates is "unified visibility." This means connecting data from different areas, including applications, infrastructure, and logs, to proactively detect anomalies in systems. The technology that makes this possible is observability. Observability involves analyzing data generated by systems to understand their internal state and identify the causes of problems. Information gathered through development and operations can also be used in security response. "In the past, companies built systems to detect and block attacks in individual areas such as infrastructure and applications. Now, they need to integrate all their data to find anomalies," Eom said. "In the AI era, the boundaries between development and operations and security are breaking down."

Datadog is expanding its security partnerships to target the South Korean security market. After appointing a head of channels and alliances in South Korea on October 26, the company plans to establish partnerships with 10 local security companies within the next one to two months. Its goal is to strengthen customer support by combining Datadog's observability capabilities with its local partners' expertise in system implementation and security monitoring. "We are building a partnership model with 10 local partners that have specialized security expertise to provide the security frameworks and solutions needed for AI governance," Eom said.

Eom identified "autonomy," in which AI assesses situations and takes action, as the future of security. He said security needs to advance from monitoring that checks for problems after they occur, to observability that analyzes data across systems to identify root causes and signs of risk, and ultimately to a stage where AI agents carry out the necessary responses. "Autonomy, with AI making decisions first and taking action before problems occur, will become increasingly important," Eom said. "Throughout this process, people must remain in control of AI agents' decisions and actions."

This content was produced with the assistance of AI translation services.