We use various systems and models, including AI-powered solutions, in analyzing and monitoring several risk categories, as well as for other business purposes. While we assess and improve these systems and models on an ongoing basis, there can be no assurance that they, along with other related controls, will effectively mitigate risk under all circumstances, or that they will adequately mitigate any risk or loss to us. As with any systems and models, there are inherent limitations because they involve techniques and judgments that cannot anticipate every economic and financial outcome in the markets in which we operate, nor can they anticipate the specifics and timing of such outcomes. Further, these systems and models may fail to quantify accurately the magnitude of the risks we face or they may not be effective against all types of risk, including risks that are unidentified or unanticipated. Our measurement methodologies rely on many assumptions and historical analyses and correlations. These assumptions may be incorrect, and the historical correlations on which we rely may not continue to be relevant. Models based on historical data sets might not be accurate predicators of future outcomes and their ability to appropriately predict future outcomes may degrade over time due to limited historical patterns, extreme or unanticipated market movements or customer behavior and liquidity, especially during severe market downturns or stress events (e.g., geopolitical events or pandemics). Consequently, the measurements that we make may not adequately capture or express the true risk profiles of our businesses or provide accurate data for other business purposes, each of which ultimately could have a negative impact on our business, financial condition and results of operations. Errors in the underlying model or model assumptions, or inadequate model assumptions, could result in unanticipated and adverse consequences, including material loss or noncompliance with regulatory requirements or expectations. In addition, the use of generative AI, a relatively new and emerging technology in the early stages of commercial use, exposes us to additional risks, such as damage to our reputation, competitive position, and business, legal and regulatory risks and additional costs.
For example, generative AI has been known to produce false or “hallucinatory” inferences or output, and certain generative AI uses machine learning and predictive analytics, which can create inaccurate, incomplete, or misleading content, unintended biases, and other discriminatory or unexpected results, errors or inadequacies, any of which may not be easily detectable. Additionally, to the extent that we do not have sufficient rights to use the data or other material or content used in or produced by the AI tools used in our business, or if we experience cybersecurity incidents in connection with our use of AI, it could adversely affect our reputation and expose us to legal liability or regulatory risk, including with respect to third-party intellectual property, privacy, data protection and cybersecurity, publicity, contractual or other rights. Further, our competitors or other third parties may incorporate AI into their products more quickly or more successfully than us, which could impair our ability to compete effectively. As the utilization of AI becomes more prevalent, we anticipate that it will continue to present new or unanticipated ethical, reputational, technical, operational, legal, competitive, and regulatory issues, among others. As a result, the challenges presented with our use of AI could adversely affect our business, financial condition, and results of operations.