Artificial intelligence (AI) is reshaping the banking sector, offering increased efficiency and innovation. However, as banks integrate AI into their operations, concerns over data privacy, regulatory compliance and ethical risks are becoming more prominent. Financial institutions must navigate these challenges while ensuring that AI-driven decision-making remains transparent and accountable.
The rapid adoption of AI has raised new questions for risk managers, regulators and industry leaders. As AI technology advances, banks must develop strategies to mitigate potential risks while maintaining a competitive edge. Alex deLaricheliere, banking subvertical leader at WTW, discussed how AI is changing the sector and the key risks financial institutions must address.
“The speed at which AI is being adopted presents challenges for technology, enterprise and operational risk managers at banks,” deLaricheliere said. While the benefits of AI – such as operational efficiencies, improved customer experiences and innovative products – are well known, the risks remain less understood.
“Many risk managers are concerned about whether their organisation’s AI usage has been adequately stress-tested.”
As AI technologies continue to evolve, it is essential for risk managers to identify and quantify future risks.
For banks, the urgency to implement AI must be weighed against the risks inherent in its rapid adoption. According to deLaricheliere, the key risks associated with AI in banking include legal and security risks, ethical concerns, and challenges related to quality and accuracy.
Data privacy and security remain a significant concern. AI systems require large amounts of sensitive customer data, increasing the risk of data breaches, hacking and misuse of information.
“While these risks are not new, AI amplifies the exposure,” deLaricheliere said. “Continuing to keep up with data privacy and compliance with stringent regulations will not be an insignificant task.”
AI also introduces intellectual property (IP) risks. Banks rely on AI to automate processes and improve decision-making, but this reliance raises questions about ownership and protection of IP, as well as potential infringement issues.
“Banks face several potential IP-related challenges, and these risks must be carefully managed to avoid legal, financial and reputational damage.”
Regulatory compliance is another area of uncertainty. “Banks are navigating a complex regulatory environment while ensuring AI technologies comply with standards for transparency, accountability and ethical use,” deLaricheliere said.
“There’s a fear of inadvertently breaching regulatory standards if AI systems don’t meet compliance requirements, especially in areas like anti-money laundering (AML) and know-your-customer (KYC).” The evolving nature of AI regulations presents an ongoing challenge for financial institutions.
Bias in AI models is a widely recognised risk. “AI can introduce bias, particularly when trained on historical data that reflects societal inequalities,” deLaricheliere said. “This could result in discriminatory outcomes, such as biased lending decisions or unequal access to financial services.”
The public’s perception of AI-driven decision-making is another concern. “Risk managers are particularly focused on how customers react when they feel that AI is making financial decisions without clear explanations.”
AI’s potential to generate misleading or inaccurate information – known as “hallucinations” – is another critical risk.
“In a highly regulated sector like banking, where accuracy and reliability are crucial, this is a particularly acute risk that can manifest itself in the form of poor financial decisions, fraud detection failures, compliance violations and a loss of customer trust,” deLaricheliere said.
Banks also face risks related to vendor reliance and automation. As they increasingly use third-party AI tools, they take on vicarious liability, which can lead to service disruptions or regulatory non-compliance.
“As banks increasingly rely on third-party vendors for AI tools and to automate various processes, they increase their vicarious liability. These liabilities can have far-reaching implications, ranging from service disruptions to regulatory and compliance issues.”
To address these risks, deLaricheliere advocated for a balanced approach that combines technology with strong risk management strategies, ethical frameworks, transparency, and regulatory alignment.
A robust governance framework is essential. “Banks must establish clear policies for AI usage, including roles and responsibilities, and create an oversight committee to monitor AI initiatives,” he said. “Strong governance ensures that AI aligns with organisational goals and is deployed responsibly.”
Mitigating bias requires continuous monitoring. “Regular bias audits of AI models are necessary to identify and address potential discrimination,” deLaricheliere said. “Using diverse datasets and employing techniques to ensure fair outcomes in decision-making processes is key.”
Transparency in AI decision-making is another priority. “Banks should strive for transparency by utilising explainable AI techniques, providing customers with clear explanations of how AI affects their financial decisions.”
From a regulatory perspective, banks must stay informed about evolving AI regulations and ensure compliance with legal requirements. “Engaging with regulators to understand expectations and collaborating on best practices will be critical,” he said.
“The absence of clear guidelines from regulatory authorities presents an ongoing challenge, so banks must remain proactive in their approach.”
AI is transforming the US banking sector by enhancing efficiencies, improving customer interactions and driving financial innovation. However, challenges related to data privacy, bias and regulatory compliance remain.
“The next few years will likely witness widespread AI adoption, reshaping the financial landscape for both institutions and consumers alike,” deLaricheliere said. “Banks must strike a balance between innovation, ethical considerations and risk management to fully realise the benefits of AI.”
What are your thoughts on this story? Please feel free to share your comments below.
