Samil PwC: Asset Management Industry Needs Governance Model for Managing AI Utilization

Report: "Governance in Asset Management Transformed by AI"

As the use of artificial intelligence (AI) expands across key work areas in the asset management industry, the existing internal control systems and responsibility structures are becoming insufficient to manage this changing environment. As a result, there is an emerging need for a new governance model capable of systematically managing all aspects of AI utilization.

On June 11, Samil PwC released a report titled "Governance in Asset Management Transformed by AI: Operational Structure Changes and Response Strategies," announcing these findings. The report analyzes the impact of AI proliferation on work operations and internal control structures in the asset management industry, and presents directions for AI governance that financial firms should establish.

According to the report, AI is now being used not only for conventional structured data analysis but also for interpreting unstructured information, drafting documents, and supporting investment decisions, thus expanding its range of applications. The spread of generative AI and large language models (LLMs) has further advanced their influence to the point of affecting decision-making processes.

This is leading to a structural shift that is reshaping decision-making frameworks, responsibility systems, and methods of internal control throughout the asset management industry. As AI-based analysis and automation become integrated into work processes, there is a growing need to restructure both human-centered operational frameworks and approval and accountability systems.

The report identifies several key governance issues the asset management industry faces as AI becomes more widespread: ▲ The spread of AI use centered on frontline work and the rise of shadow AI (use outside of organizational controls); ▲ Increasing complexity of information flows and greater management burdens due to expanded use of AI-based data; ▲ The need for complex management in response to a multi-layered regulatory environment; ▲ The emergence of new security threats such as data leakage and prompt injection due to the characteristics of generative AI; ▲ And greater supply chain and external dependency risks arising from the use of cloud, API, and external AI services.

In particular, the growing use of shadow AI without internal approval and increased reliance on external generative AI services represents a new risk that existing internal control systems have not accounted for. Managing input information and clearly establishing boundaries of responsibility have emerged as major challenges in this context.

The report highlights "data governance and human-centered decision-making frameworks" as the core pillars of AI governance. As the expansion of AI use involves integrating external data into AI services, it is vital to maintain a data-centric control system that consistently manages data sources, scopes of use, and movement paths. The report also emphasizes that for areas such as investment decisions and client-related judgments, organizations must ensure accountability by having human review and approval procedures, rather than relying solely on AI-generated results.

To address these risks, the report calls for company-wide governance measures such as: ▲ Establishing AI use policies and operational standards; ▲ Building data governance focused on the management of data sources, quality, and records; ▲ Strengthening decision-making frameworks based on human review; ▲ Setting up AI usage logging, monitoring, and audit response systems; ▲ Upgrading security and access control frameworks; ▲ And improving management of external AI services and supply chains. Since AI utilization is subject to a complex array of compliance and risk requirements—such as the Personal Information Protection Act, the Credit Information Act, and electronic financial regulations—organizations must develop integrated management systems at the organizational level.

Haemin Jung, AX Node Partner at Samil PwC, stated, "The future competitiveness of financial firms will not be determined by AI utilization itself, but by how responsibly and reliably they can manage it—in other words, by their AI governance capabilities. While AI is a powerful tool for enhancing information analysis and work efficiency in asset management, it also presents new challenges for existing operational frameworks and internal control systems, making the establishment of integrated company-wide management systems an urgent priority."

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