The financial services industry has a history of using quantitative methods and algorithms to support decision making. These are a foundation of AI systems, and the industry is therefore primed for AI adoption, positioning it at the forefront of adopting and benefiting from AI technologies, according to a report from UK Finance and Microsoft.
AI can build on human intelligence by recognizing patterns and anomalies in large amounts of data, which is key in applications such as anomaly detection (e.g. fraudulent transactions). AI can also scale and automate repetitive tasks in a more predictable way – including complex calculations, for example for determining risk.
One of the most important types of AI is machine learning (ML), algorithmic systems that can recognize patterns and learn without being explicitly programmed. ML has become one of the key AI technologies used by the financial services industry, due to its ability to use existing algorithms to improve ever-growing amounts of data, thereby creating new capabilities.
AI will become more and more crucial as the data that we create continues to grow. This will lead to a situation where processes that previously did not require AI (e.g. fraud detection), will no longer be able to succeed without it.
As more complex use cases are built with AI, thought and design is needed in the explainability of the decisions or outputs produced by these tools. Full explanations may be needed to reassure customers as to why their credit or mortgage application was not accepted by an AI/ML model. Where firms identify a trade-off between the level of explainability and accuracy, firms will need to consider customer outcomes carefully. Explainabilty of AI/ML is vital for customer reassurance and increasingly it is required by regulators.
“We have a set of risks around data privacy and security that are forefront. We need to be more aware of consequences, and appropriate steps should be taken,” said the head of Business Design at a top tier UK bank, cited in the report.