This is the second in a series of articles on the challenges, limitations, and opportunities presented by artificial intelligence (AI). In our first article, Rise of the Machines?, we looked at the application of AI in our own due diligence industry. We acknowledged the transformative effect it has on data gathering but guarded against any notion that its ready to start competing against the innate intuition of a seasoned professional when it comes to the analysis of that data. This time we turn our attention more broadly to the use of AI within the financial sector, looking at the extent of its penetration, the different dimensions of regulatory and compliance risk while standing these alongside the upside benefits ai brings.
There’s little doubt that financial institutions have been among the leading early adopters of AI technology and that sentiment on its value remains strongly bullish. But this is increasingly being tempered with a realistic view of the extent to which AI has replaced human activity. Speaking in a recent edition of Goldman Sach’s Global Macro Research, Daron Acemoglu, Institute Professor at MIT, estimates that only a quarter of AI-exposed tasks will be cost-effective to automate within the next 10 years — which means that AI will impact less than 5% of all tasks and boost US productivity by only 0.5% and GDP growth by 0.9% cumulatively over the next decade. “Truly transformative changes won’t happen quickly, and few will likely occur within the next 10 years,” Acemoglu says.
AI Technologies – Regulators & Compliance
AI technologies in the financial sector are subject to a complex web of regulations that vary across jurisdictions. Financial institutions must comply with regulations covering data privacy, consumer protection, anti-money laundering (AML), and anti-discrimination laws. The use of AI can complicate this compliance in several ways, most notably:
- Data Privacy and Security: AI systems require vast amounts of data to function effectively. This often includes sensitive personal and financial information. Regulations like the General Data Protection Regulation (GDPR) in Europe and the California Consumer Privacy Act (CCPA) in the United States impose strict requirements on how data is collected, stored, processed, and shared. Non-compliance can result in hefty fines and reputational damage.
- Anti-Money Laundering (AML): AI is increasingly used to detect suspicious transactions indicative of money laundering. However, if these systems are not properly calibrated or lack transparency, they can either miss illicit activities or generate false positives, both leading to compliance failures and potential penalties.
AI systems, particularly those using machine learning, can inadvertently perpetuate or even exacerbate biases present in the data they are trained on. This can result in discriminatory practices in areas such as credit scoring, loan approvals, and customer service. Discriminatory practices not only violate ethical standards but also contravene laws such as the Equal Credit Opportunity Act (ECOA) in the U.S., which prohibits credit discrimination based on race, color, religion, national origin, sex, marital status, or age. Similarly, models, especially those based on complex algorithms like deep learning, can be opaque and difficult to interpret—often referred to as “black box” models. This lack of transparency poses a significant compliance risk, particularly in scenarios where banks are called upon to explain decisions, such as loan rejections or pricing.
While there are inevitably jurisdictional variances on these compliance matters, its notable that regulators seem to be taking a more multilateral approach to how they view the potential effect AI may have on market competition. On July 23, 2024, the competition authorities of the EU, the UK, and the U.S. issued a joint statement on competition in artificial intelligence foundation models and AI products. Since the emergence of generative AI, each of the authorities has individually been studying the potential risks to competition and while the Statement represents only a high-level summary of their collective concerns at this point in the development it nevertheless signals their collective intent to treat the area as an enforcement priority.
AI Challenges for IT Infrastructure
There are challenges also for financial institutions dealing with dated and/or disparate information technology infrastructure. A recent KPMG study found that 40% of financial services executives surveyed said their organization’s IT and digital infrastructure would limit their ability to innovate and adapt to advancements in AI. Financial firms built on legacy systems come with a price (for maintenance), leaving less money on the table to pursue new technologies, the survey found. “Banks are moving more work to cloud environments, where they would have more flexibility and agility in adding new capabilities such as LLM models, but the process is slow,” the KPMG report says. “Banks can also face organizational barriers that limit their agility versus fintech startups. As a result, they may struggle to keep pace with the rapid innovation and product development cycles of cloud-native and AI-native startups, losing market share to these more agile competitors. Legacy asset managers face some of these same challenges.”
While its certainly true that an over reliance on AI can generate risk where institutions do not fully understand the limitations and appropriate use cases, it’s also true that the opportunities remain significant. George Lee, Partner and Co-Head of Goldman Sachs Global Institute is quoted in Business Insider magazine as noting “it could take years for the consequences of this tech (AI) on the bank’s workflows to be fully understood. It is going to elevate the creativity and creative problem-solving that our bankers bring to clients.”
AI Brings Significant Efficiencies to the Venture Capital Space
The venture capital space is embracing AI, if nothing else for survival’s sake. “Firms that do not use AI to source deals will be left behind,” says Sri Chandrasekar, Managing Partner, Point72 Ventures. Consequently, significant changes are taking place in the VC space at a rapid pace with some estimating that back offices could be reduced by as much as 50 percent. This same Business Insider piece cites a prior “much discussed column last year” from James Currier, General Partner, NFX, who commented on why he thinks the use of AI will level the playing field for venture investors in the next decade, like what software did for stocks and bonds in the past 40 years. Currier, however, remains a believer that analysts and senior relationship managers will remain vital as he states, “Venture firms will have to remake themselves into a combination of people and AI.” Chandrasekar echoes this hybrid approach as the likely result in the future, “At its heart, venture is not about finding the best deals but winning them. That comes down to a personal connection between founder and investor.”
The benefits of incorporating AI into processes currently in place at financial institutions are thought by some to be quite positive not only in production but also to the bottom line. According to a recent study by consulting firm Deloitte, generative AI will unleash “a new era of productivity” for investment banking, boosting the productivity of front-office workers by as much as 35% by 2026, which could translate to additional revenue of at least $3 million per employee.
A Balanced Approach Yields Winning Formula
Others working closely with Wall Street firms and particularly their new hires also believe the combination of technology and human expertise will be the winning formula for financial institutions. Bogdan Tudose of Training the Street (TTS), which preps up-and-coming finance and business students and new hires at client firms like Blackstone, JPMorgan, and Morgan Stanley, warned against buying too heavily into doomsday predictions. “Generative AI ‘hallucination’ is a real problem right now,” he added, comparing it to the introduction of Excel, which had people predicting that accountants and financial analysts would be out of a job. “Sure, it replaced some very low-entry jobs, but overall models have become more complex and powerful,” Tudose said. “Tools such as generative AI will have the same impact. They will help automate some of the mundane, repetitive tasks so that analysts can focus on more value-add analysis and apply critical thinking to their models.”
AI in the Financial Sector has far reaching possibilities and a balanced approach will drive success. Integrity Risk International CEO, Jim McWeeney, provided his summation: “For us, adding client value by freeing talent from the drudgery of mundane tasks to focus on a higher level of analysis and critical thinking continues to be where the real power of AI lies.”
