This ISDA Future Leaders in Derivatives (IFLD) whitepaper sets out guidance for industry stakeholders, regulators and technology providers seeking to harness the power of generative artificial intelligence (genAI) in transforming the over-the-counter (OTC) derivatives market.
It discusses promising use cases for this technology in the OTC derivatives market, the risks of such technology, the current regulatory framework and potential risk mitigation solutions. By embracing genAI, organizations can unlock new opportunities for efficiency, innovation and compliance in an increasingly complex market.
GenAI is rapidly becoming a useful tool across industries, with the ability to significantly reduce the time taken to produce work. There are several promising use cases for genAI in the derivatives market. The first relates to its ability to create new language based on precedent and synthesize data into a human-readable summary. GenAI is a useful tool for market participants to summarize complex derivatives agreements and suggest clauses based on deal terms and firms’ existing precedent agreements, which has the potential to significantly decrease negotiation and drafting costs. In addition, genAI can be used to extract unstructured data from derivatives documentation to provide summaries of derivatives transactions required for operations and front-office processes.
GenAI can also synthesize various jurisdictional regulations and present these in an easy-to-read format, comply with industry or firm standards and provide checks against trades and trade documentation. While not a replacement for a human lawyer, it can significantly accelerate the review process and act as an additional regulatory compliance check.
The second use case identified is with respect to genAI’s use in application development to propose new code changes. McKinsey has estimated that using genAI in this way can make coding up to 56% faster. The third use case is to analyze data, including nuanced human emotion data, to provide market insights that can be useful in trading. The fourth use case is to improve operational efficiencies, such as to summarize margin and collateral requirements for the business and assist in selecting the least costly collateral or create synthetic data that can be used for model testing. Finally, genAI can be used to assist in the development of derivatives markets in emerging markets, by aiding firms in summarizing local regulations and market conditions, paving the way for a more efficient entry into such markets.
Governments are also looking at use cases for genAI and proposing regulations to safeguard consumers and financial markets. These proposals are still in their infancy, but reviewing the current state shows the direction in which regulators and policymakers are heading.
While these use cases offer great efficiencies, the use of genAI does not come without its challenges and risks. Due to the nature of genAI and the large amount of data needed to train the models, data breaches can be a significant challenge and lead to reputational, confidentiality, intellectual property and legal risks. The use of genAI for trading can also create regulatory issues and, without proper oversight, could lead to fines and sanctions from financial regulators. Additionally, genAI is associated with producing bias and could be used to discriminate against protected classes, leading to civil and possible criminal liability for companies. Lastly, there is significant risk of model failure, in which the results produced are sub-standard or simply false. This could lead to erroneous trades and diminished trust within a financial institution.
There are steps companies can take to mitigate these challenges and risks. This paper concludes with a set of proposed best practices. Firms are encouraged to formulate a comprehensive governance framework for their own use of genAI and that of their third-party vendors. In addition, companies should ensure the recommendations and presentation of material produced by AI are acted upon only by humans who can take responsibility for the ultimate decisions of the company. Firms should implement comprehensive cyber security and data security policies to safeguard their IT systems from cyberattacks and malicious use of genAI technology. Lastly, firms should develop model risk mitigation policies to ensure the models conform to expected results and any deviation or failure can be quickly corrected.