2022 Top Predictions for AI in Finance

What’s likely to occur in artificial intelligence in the world of finance in 2022? Here’s what leading academics, analysts and AI experts had to say: 

Increases in Using Synthetic Data Sets

Lukasz Szpruch, associate professor at the University of Edinburgh’s School of Mathematics and program director of the Finance and Economics Programme at The Alan Turing Institute

“Whatever you are trying to do using data-driven tools, the data is at the core of it. We’ve learned that data is not perfect and that biases exist. The challenge with cases like fraud detection is that each financial institution is only seeing the world from the lens of its own data sets. So much more could be done if we were able to bring data from across other institutions and be able to track those malicious actors. This is an old idea that’s being reheated as we can now do it better. The same idea is being used under the name of market generators in more quantitative financing, where we are now beginning to be able to automate pricing derivatives.”

Finance Will be Quantum’s First Serious Test

Alexander Harrowell, senior AI and IoT analyst at Omdia

“It’s still to be seen whether quantum computing will be genuinely useful and how quickly. Financial services customers have been some of the first to trial the technology in production, helped by a strong fit between problem sets such as portfolio optimization and the kind of binary optimizations that current quantum systems do well. Over the next two years, Omdia’s Quantum Computing: State of the Market Survey suggests we can expect a five times increase in projects in production, a 7.5x increase in pilot projects and the first scale-up projects. Many of these will be financial. This also means that financial users will be the first to encounter the problems. Half of our respondents said their biggest barrier to adoption was ‘no understanding’ of what the technology can do. Over the next two years, they will be on point in finding out.”

Ethical Frameworks Becoming More Common

Kai Yang, chief data officer, APA at HSBC

“We expect to see ethical AI frameworks becoming a more common feature of responsible corporate governance, as regulators take a stronger and more active stance on the fairness of banking processes and models. Customers too are demanding greater levels of transparency around how their data is used; hence, ethical AI culture will need to become an integral part of corporate identity. At HSBC we aim to take a leadership role, as one of the first financial service companies to create AI and data ethics principles, and recently partnering with the Monetary Authority of Singapore and the Alan Turing Institute to help develop a framework for responsible adoption of AI in the financial services industry.”

Humans and AI Will Increasingly Interact in a Seamless Integration

Manuela Veloso, head of AI research, JPMorgan Chase

“Through language and image processing and machine learning, AI will enable at large scale, the search and understanding of the never-decreasing available digital data. AI will help with data standardization, pattern detection, safe data sharing, prediction and anticipation. As we face increasingly complex decision making involving many participants and many objectives, we will rely on AI assistants to tediously analyze, simulate and evaluate large numbers of alternative solutions. Humans and AI will increasingly interact in a seamless integration of their capabilities in a continuous learning experience. AI systems will include explanations and actively request data and feedback to improve their assistance over time, with the goal to capture underlying human values and rules. Overall, we will continue to experience AI enabling human dreams to improve life in all sorts of ways, including health, finance, climate, energy, education, equality and social condition.”

Expect Incoming Regulation

Felix Hoddinott, chief analytics officer, Quantexa

“Historically, the complexity of deploying AI models for regulatory purposes has blocked AI initiatives within many financial institutions. But regulators are increasingly seeing evidence of the impactful improvements achievable from using AI applied to the wider data describing the full context around decisions. Regulators will now issue guidance to accelerate this use of AI, especially in areas like risk assessment and monitoring. This will not reduce the requirements for justifiable and fair models but clarified guidelines will be more clearly open to addressing these requirements through emerging technologies and methods. Establishing modern governance processes to simplify deployment in a regulatory space will reduce risk and improve customer experience.”

Increased Personalization of Finance Products

Farouk Ferchichi chief data analytics officer, Envestnet 

“In 2022, AI will be harnessed in finance to create a hyper-personalized and unified customer experience, to reduce costs, and to target offers and cross-sell products. Given ongoing regulatory pressure, financial companies will utilize AI to improve and automate the monitoring of data quality, especially for product data that is used for regulatory reporting. In addition, the scope of model governance will continue to expand, with financial institutions having to rely on a combination of synthetic data to test models, as well as alternate data as a backup. AI can enable financial firms to segment product offers by market audience, and distribute them as part of an integrated, hyper-personalized omnichannel experience for customers.”

Almost All Finance Firms to Create Shared Threat Intelligence Organization

Helen Sutton, SVP EMEA and APAC sales, Dataminr

“We’re seeing a growth in which financial services institutions (FSIs) are looking to implement more thoughtful digitalization. With that comes increased risk. In fact, over 700 organizations experienced a ransomware attack in Q2 of 2021 and the average ransomware pay-out has almost tripled what it was last year, with organizations paying $850,000 on average. I foresee that almost all FSI’s will review in 2022, if not by Q4 of 2021, whether to create a shared threat intelligence organization between cyber and physical threats. More than ever, banks and insurers need to de-silo their critical information structures to ensure effective support and security when adopting new technologies and platforms. We’ll see further investment in the applications of AI that support this parallel trend.”

As Deployments Increase, Expect Fears to Dissipate

Mike de Vere, CEO, Zest AI

“We predict that AI will continue to push its way into more critical functions within the financial services industry. For example, we’re seeing AI-driven credit underwriting become more popular in unexpected places such as smaller regional lenders and credit unions. There’s an enormous data arbitrage to be gained by replacing legacy FICO scoring with AI-based models. We’re talking upwards of 30% to 50% statistical improvement for hard-to-score consumers, which translates into hundreds of millions of dollars in profit for lenders. As more AI is integrated into businesses, we’ll see fears around its use subside. Employees will become more comfortable with the technology and realize the potential it has to improve the overall quality of work their teams produce. The fears about the technology replacing employees will ultimately shift into an appreciation for the technology’s capabilities.”

AI to Play a Larger Role in Risk Assessment

Kenneth Chan, managing director and co-founder, ViewTrade Holding Corp.

“It is no secret that AI has played a major role in the ongoing democratization of investing. My prediction for next year and beyond is that the major growth we’ve seen in retail investing will continue at a rapid pace – and AI will continue to fuel that growth. AI has helped to level the playing field for investors. Today you don’t have to be a high-net-worth (HNW) investor to get personalized financial advice, there is a chatbot for that. These AI-driven chatbots will only continue to get smarter. Machine learning can now sift through various financial accounts and profiles for a user and provide a snapshot of recommended to-dos on a dashboard. This will continue to gain traction in the decade ahead. AI has also helped to simplify the client onboarding process, while also enhancing the customer experience. Going forward, as the retail investing trend continues to grow expect AI to play a larger role in risk assessment, risk management, and fraud detection. This will enable businesses to scale and keep up with heavy volatility.”

This article first appeared in IoT World Today’s sister publication AI Business.