Finance is a numbers game. It’s increasingly an AI-powered one too.
“One of the most significant capabilities of AI is its ability to turn massive volumes of data into actionable information. This makes AI naturally fit the financial services industry, where data is the core asset and underlies virtually all business activities,” says Haoran Wu, assistant professor in banking and finance at UCD.
As a result, the financial services world is undergoing a transition from digital finance to AI-driven intelligent finance.
AI has been increasingly integrated across the financial services sector to improve operational efficiency, decision accuracy, risk management and customer service, he points out.
READ MORE
In banking services, AI is adopted for credit scoring and loan approvals based on customers’ financial and behavioural information.
“Banks are also employing machine-learning algorithms, a subset of AI, that learn from vast amounts of transaction data to detect money laundering and protect customers from fraud,” says Wu.
In investing, AI analysis tools are increasingly integrated into fintech platforms, such as stock-trading apps. For wealth management, financial institutions are deploying AI-powered robo-advisers to deliver personalised investment recommendations.
In customer service, “generative AI-powered conversational systems are progressively replacing traditional keyword-based chatbots, allowing more natural and context-aware interactions,” says Wu.
But despite the evident benefits of AI adoption, there are growing concerns about whether AI is used ethically and responsibly in financial services.
“Consumers are concerned about data privacy and the potential use of their personal information to train AI models. Moreover, because AI models are trained on historical data, they may inadvertently inherit existing biases and discrimination in decision making, resulting in algorithmic discrimination.”
Another big concern is the lack of transparency of AI systems, which makes it difficult for consumers to understand how decisions are generated.
To counter these risks, various regulatory frameworks for AI have been mandated worldwide. The EU AI Act, enforced in August 2024, classifies AI according to its risk and imposes corresponding obligations on different categories of applications.
“Financial institutions deploying these technologies are therefore required to comply with stringent obligations related to transparency, human oversight, data governance and detailed record keeping,” says Wu.
AI is now a key enabler of the ongoing transformation of global financial services, says Darren O’Neill, insurance partner at PwC Ireland.
“As well as bringing operational efficiencies, enhancing customer experience and managing risks, it includes the intelligent automation of complex processes using AI agents, the improvement of fraud detection by analysing larger structured and unstructured data sets, and the use of natural language processing to power the chatbots, robo-advisers and virtual assistants.”
Large language models (LLMs) underpin this shift in how AI is being more routinely deployed.
“LLMs have the ability to process and understand vast amounts of complex, unstructured data quickly and accurately. They help in automating report generation, summarising market trends and extracting insights from vast unstructured data like news, earnings calls, and regulatory documents,” says O’Neill.
“They assist in risk management by analysing textual data for early warning signals. Additionally, LLMs improve compliance monitoring by interpreting regulatory language and detecting potential violations, thereby increasing operational efficiency and accuracy in financial institutions.”
It’s why “responsible AI” is now an executive- and board-level priority for organisations, he points out.
“It refers to the critical requirement to build trust, ensure fairness and mitigate risks in AI deployment,” he explains.
“Many organisations have established governance frameworks that underpin the continuous monitoring of deployed AI to prevent unintended consequences,” says O’Neill.
“While this is good practice and aligned with the regulatory requirement, many organisations simply understand the criticality in fostering consumer confidence by being transparent in how they are using their data.”
Ensuring consumer confidence is critical.
“Financial institutions are implementing governance, risk management and compliance frameworks to ensure AI systems operate safely, transparently and within regulatory requirements,” says Rory Timlin, partner, management consulting, at KPMG.
“Banks that are deploying agents into the flow of work in their organisations are developing monitoring and tracking frameworks to help ensure that agent behaviour is appropriately implemented, measured and controlled.”
There is also a growing recognition that, in terms of cybersecurity, AI is both a threat and an opportunity. “AI is increasingly featuring as a mechanism creating new cyberattacks and risk, while also being used internally by companies as a capability to increase cyber resilience,” says Timlin.
“The ECB has recently issued instructions to banks in relation to addressing AI-enabled cyber threats, such as Mythos, with action plans to be submitted in the autumn.”
As new AI-powered businesses emerge, consumer confidence will be critical.
“A mortgage can determine where someone lives and how a family plans its future. So the right question is not whether AI should decide who gets a mortgage. It is where AI can remove friction while keeping advice, underwriting and accountability with people,” says Jamie Lawless, chief executive of Irish AI-native fintech LendWell.
“That is the problem LendWell was built to solve. The platform uses AI to collect, read and classify documents, extract the relevant facts, flag missing information and track a mortgage case from first inquiry through to completion. The aim is not for AI to say yes or no. It is for every adviser and underwriter to see what is complete, what is missing, what has changed and where each piece of evidence came from.”
Finance Solutions, a mortgage broker, has partnered with the AI-powered platform to enhance customer service.
“Responsible AI in lending should be judged by how it can best support a business and its customers,” says its managing director, Conor McGowan.
“Our people advise and support. We are using AI to automate and improve parts of the administrative process involved in securing a mortgage, and, used properly, AI will make mortgages faster, clearer and more human. Not less.”
