Financial authorities accelerate AI use guidelines in line with Lee's initiative

gettyimagesbank

Financial authorities are accelerating efforts to revise guidelines on the use of artificial intelligence (AI) in the financial sector, in line with the newly launched Lee Jae-myung administration's push to make AI-related legislation a core policy, officials said Thursday.

Lee has pledged to invest 100 trillion won ($73 billion) in AI and to position Korea as one of the world's top three AI powerhouses. He has also established a senior presidential secretary for AI within the presidential office.

To align with the new administration's AI policy direction, the Financial Services Commission (FSC) recently signed an additional contract worth 20 million won to broaden an ongoing research project aimed at revising AI use guidelines in the financial sector.

Initially launched in March, the study was extended to incorporate recent shifts in AI-related policies and to deepen its scope and focus.

The existing guidelines, introduced in December 2021, are nonbinding recommendations that urge financial institutions to uphold principles such as fairness and accountability when deploying AI.

However, critics have long argued that these guidelines are overly vague and lack legal force, making them insufficient for keeping up with the fast-evolving technology landscape.

The country's top financial regulator had formally announced its intention to revise the guidelines in December. However, political instability at the time, sparked by ousted President Yoon Suk Yeol's martial law declaration, delayed the start of the work.

In the meantime, generative AI gained traction rapidly in various financial services such as loan evaluations, insurance underwriting and customer support, further exposing a regulatory vacuum.

gettyimagesbank

One of the major concerns surrounding the use of AI in finance is the ambiguity around accountability when mistakes occur in AI-driven outcomes.

For example, if a generative AI system provides flawed financial advice, it remains unclear whether the financial institution or the AI developer should be held responsible.

As a result, one of the central issues in the FSC's ongoing revision effort is the development of a governance framework to manage the use of generative AI.

Critics have also pointed out the lack of effective measures to uphold consumers' rights, such as the right to receive an explanation or to challenge decisions made by AI systems.

Although the existing guidelines advise financial institutions to provide explanations that are understandable to consumers, many firms have in practice avoided doing so, often citing the technical complexity of their systems or limitations of their AI models.

The research project is scheduled to be completed by next month. The FSC plans to then prepare a draft for revised guidelines in the second half of the year.

"Once the draft revision is ready, we will actively gather industry feedback and conduct pilot implementations to establish a regulatory framework that can be applied in practice effectively," an FSC official said.