A number of factors have contributed to the rise in adoption of artificial intelligence (AI) in the financial sector in recent years. These include improvements in deep learning, data collection, cloud computing, advances in software as well as hardware developments and so on.

Although there is not enough data to carry out a more detailed use case analysis, a recent report by the Financial Stability Board (FSB) has found that AI adoption in the financial sector is on the rise. The report also suggests that most use cases of AI in the finance sector revolve around two key areas; enhancing internal operations and improving regulatory compliance.

The FSB had published a similar report back in 2017. This latest work revisits the 2017 findings, but also looks int the latest advancements in AI technology, and its adoption by the financial sector.

At present the use of AI to generate new revenue streams in the financial sector is not widely observed. However, this could change in the future as the technology finds its way into other sectors within the financial industry.

Similar findings

These findings reflect similar results from a recent study by the Boston Consulting Group which looked into how firms around the world were finding value in AI adoption.

According to that research, only 4% of the more than 1,000 global firms who took part are utilising AI to find true value, be it in revenue generation or other sectors of their businesses.

According to the FSB report, AI technology is developing at a rapid speed, especially generative AI (GenAI). The financial industry is trying to keep up with the advances, though with some caution around the use of GenAI. At the same time, regulators are also adopting AI in order to better perform their responsibilities.

But there are potential challenges and risks. “The fast pace of innovation and AI integration in financial services, along with limited data on AI usage, poses challenges for monitoring vulnerabilities and potential financial stability implications,” the report warns.

Vulnerabilities

The report highlights a number of ways in which rapid and large-scale AI adoption can also leave the financial sector vulnerable to risks and complications.

• AI adoption brings with it a certain degree of dependency on third-party service providers. This means the use of specialized hardware, and pre-trained models. Any disruption to these service providers could also potentially expose the financial sector to operational disruptions and systemic risks.
• A large chunk of the financial sector could also be using the same AI models and data sources on a daily basis. This could mean increased correlation in trading, lending and pricing. “This could amplify market stress, exacerbate liquidity crunches, and increase asset price vulnerabilities,” says the report.
• Increased AI adoption and increased dependency on specialized third-party service providers could also expose the financial sector to more cyber threats, especially from actors adept in using AI technology.
• Most AI models are complex and the quality of the data they use is hard to assess. This exposes the financial sector to model risks because of a lack of comprehensive AI governance.
• The use of technology, especially GenAI, by malicious actors for the purpose of spreading misinformation and confusion could also disrupt financial market stability.
• Lastly, “and from a longer-term perspective, AI uptake could drive changes in market structure, macroeconomic conditions, and energy use that, under certain circumstances, could have implications for financial markets and institutions.”

Use cases

According to the report, financial institutions as well as regulators are using Ai in a number of ways, and for varying purposes.

• Customer-facing: The use of AI models in credit underwriting, insurance pricing, client-facing chatbots and marketing.
• Operations-focused use cases: This includes capital optimization, model risk management, market impact analysis, and code generation.
• Trading and portfolio management: To assess market sentiment from text data, such as earnings calls or regulatory disclosures, or to implement reinforcement learning for trade execution.
• Regulatory compliance: To facilitate investigations into sanctions evasion, to identify misuse of legal persons and legal arrangements, to uncover trade fraud and trade-based money laundering, and to detect tax evasion, fraud/scams, and money mules.
• Regulatory and supervisory use cases: Financial sector authorities are also engaging with AI through a variety of use cases, including the use of technology by supervisors (for example SupTech).44 Supervisory authorities’ use of SupTech has increased, with 59% of authorities surveyed using various applications in 2023,

Steps to minimise risks

In its conclusions the report also suggests a number steps that financial institutions and national authorities can take in order to minimise the potential risks arising from the rapid and large-scale adoption of AI technologies within the sector. These include:

• The sector should closely monitor developments in AI use and its implications on financial stability. There is a need to address data and information gaps in their monitoring strategy.
• Institutions should take a look at the potential vulnerabilities highlighted in the FSB report, and then make an assessment of whether their regional/ national regulatory and supervisory frameworks provide enough protection against those vulnerabilities and risks.
• There is need for more international and cross-sectional sharing of information and good practices around AI adoption in the financial sector. This could help enhance the regulatory and supervisory frameworks at different levels and minimise potential risks.

Overall, the report suggests that “since 2017, the adoption of AI tools in the financial services industry has not only become more widespread but the use cases have also diversified.”

The benefits of AI adoption by the financial industry are numerous. But, at the same time, such a rapid and, at times, uncontrolled adoption could also “amplify certain financial vulnerabilities with potential implications for financial stability.”