Rohan Whitehead from the Institute of Analytics (IoA), walks us through how AI revolutionises the finance industry
The Institute of Analytics (IoA) has a strong interest in AI and actively engages with the topic. While AI brings undeniable benefits to various sectors, it is important to acknowledge the risks that arise from unregulated technologies.
AI has received a surge of interest in the finance industry, revolutionising various aspects of the sector. This shift has been facilitated by the abundance of data, including numerical datasets and online text-based information.
As organisations adopt hybrid cloud solutions and increase their computational capabilities, they can maintain high performance at scale, giving them a competitive advantage. AI implementation empowers financial institutions to make data-driven decisions, gain valuable market insights and enhance overall performance.
AI has diverse applications within financial institutions, encompassing fraud detection, multi-agent trading simulations, customer experience improvements, understanding client intent, interpretation of legal language, and the automation of policy enforcement.
It’s important to note that while data is a fundamental component of AI, it can also be a customer of AI. Not all proprietary data is suitable for model training, and some businesses lack a sufficient sample size to develop their own tools. AI can assist in proxy/synthetic data creation, stimulating further growth and contributing to developing more robust AI algorithms.
AI applications within the finance industry
To examine real-world examples of AI’s use cases in finance, we can look at JP Morgan, the bank with the highest market cap, and value of total assets, excluding Chinese banks. JP Morgan has publicly prioritised an “AI-first” approach and has undergone various business transformations to break down legacy silos and create horizontal data consolidation and collaboration platforms.
In 2021, following the volatility in retail stocks like GameStop, traders recognised the disproportionate impact of retail activity on certain stocks. They sought a structured approach to monitor social conversations for predicting short squeezes and implementing timely risk mitigation measures.
This scenario exemplified the need for a proactive “leading indicator” with strong predictive capabilities rather than relying solely on retrospective indicators. By considering historical data and integrating the concepts of influence and virality into their social analysis model, the JPM team developed a leading risk mitigation algorithm, which earned them accolades.
Tasks that require multiple search processes
Furthermore, tasks that require multiple search processes can be tedious and lengthy in the workplace, with significant levels of variability in completion time based on the employee’s efficiency. The problem is that, till recently, search engines are not task-aware and hence cannot help with the objective of the search beyond documentation and information retrieval. However, companies like JP Morgan are implementing auto-responsive proactive models that constantly monitor business processes and understand task context.
The accessibility of big data, relevant to business use cases has opened up numerous opportunities for AI within the finance industry. For example, asset management benefits from leveraging historical data, RSS feeds and emotional analysis to identify market trends and investment prospects. Alternative data sources such as satellite imagery and consumer sentiment further enhance the ability of algorithms to identify crucial data points, facilitating informed investment decisions.
Computing power affordability
The affordability of computing power plays a pivotal role in processing and analysing massive volumes of real-time data. Advancements in cloud computing have enabled AI algorithms to operate efficiently on a larger scale. A survey by NVIDIA in February 2023 revealed that almost half of the surveyed firms were transitioning to a hybrid cloud model, cost-effectively accessing powerful computational resources.
This has particularly benefited algorithmic trading, where real-time market decisions require swift analysis and execution, and algorithms need to respond to new data on the fly. Integrating diverse datasets and computational capacity allows financial institutions to make time-efficient, data-driven decisions, giving those with greater resources a significant competitive edge.
Consistent ethical use of AI in the future
As new technologies rapidly advance, they risk being exploited while regulators try to catch up. While AI brings numerous benefits to the financial sector, it is crucial to prioritise its safe and ethical implementation. Taking necessary measures now will establish a foundation for ensuring consistent ethical use of AI in the future.
Regarding financial consumer and investor protection, AI applications in finance can indeed create or intensify non-financial risks. Using algorithms in financial decision-making may lead to unintended biases or discriminatory practices, posing risks to consumers and investors.
How policymakers should consider AI
Policymakers should examine the implications of these technologies and carefully consider the benefits and risks associated with their use. By developing regulatory frameworks, policymakers can ensure consumer protection and ethical use of AI in the financial industry.
With great innovative leaps comes the risk of exploitation of new technologies as regulators scramble to play catch up. Recently, the European Parliament passed the Artificial Intelligence Act, a significant step towards adopting comprehensive rules for regulating AI technology.
The act categorises AI into four risk levels, with systems considered an unacceptable risk, such as “social scoring” and predictive policing tools, being banned. AI applications involving vulnerable populations and hiring practices would face stricter scrutiny. The regulations also provide provisions for privacy standards, transparency and impose substantial fines for non-compliance.
The passing of the AI Act by the European Parliament demonstrates policymakers’ recognition of the need to regulate AI technology. The act aims to establish comprehensive rules to mitigate risks and safeguard consumer investor interests. However, considering the rapid progress of AI technology, policymakers face challenges in keeping up with its advancements.
The AI Act represents a significant opportunity for the EU to establish lasting guardrails for AI, emphasising the need for the regulations to stand the test of time, with hopes that other parts of the world will follow suit through international collaboration.
