FSB highlights importance of data quality in AI for financial stability
US Treasury Official Nellie Liang, Chair of the Financial Stability Board (FSB) Standing Committee on Assessment of Vulnerabilities, addressed the OECD-FSB Roundtable on Artificial Intelligence in Finance. Liang emphasised the transformative potential of AI in finance, however she also highlighted the need for robust data quality to manage the risks associated with AI deployment.
Liang outlined how financial institutions are increasingly leveraging AI for cost reduction, risk management, and service customisation. However, these benefits come with substantial risks, particularly related to model accuracy and data integrity.
The reliability of AI outputs depends heavily on the quality of the underlying data. Inaccurate or biased data can lead to flawed decision-making, which is particularly perilous in financial contexts. Liang stressed that managing data quality is essential to ensuring AI models are effective and safe, echoing the necessity for stringent governance and oversight frameworks.
From a regulatory perspective, Liang advocated for the enhancement of existing frameworks to address the unique challenges posed by AI. She suggested that regulators focus on amplifying known risks, such as data quality issues, and adapting oversight mechanisms to cater to new AI-driven risks. Ensuring transparency and accountability in AI applications remains paramount to safeguarding financial stability.
XBRL International recognises the crucial role of high-quality, structured data in harnessing the full potential of AI while mitigating its risks. As AI continues to evolve, maintaining rigorous data standards and regulatory oversight will be vital to ensuring the stability and integrity of the global financial system.
For Nellie Liang’s full remarks, see here.
