Financial services have not been successful in adopting artificial intelligence, European fintech executives claim, despite growing evidence that the much-hyped technology boosts productivity and cuts costs..
Fear of job loss, regulatory concerns, and institutional inertia are factors that prevent bankers from fully embracing the systems that underpin products like ChatGPT.
“Big banks will never hire you. [the technology] “It won't be as fast as other fintech companies,” says Tom Bromfield, a group partner at Silicon Valley startup incubator Y Combinator and co-founder of Monzo, but generative AI will “make banks more efficient and allow them to offer the same products at a lower cost.”
According to Capgemini research, only 6% of retail banks are ready to deploy AI at scale across their operations, but McKinsey estimates that AI could add up to $340 billion in value to the global banking sector annually, equivalent to about 4.7% of the industry's total revenue.
With the ability to answer questions and analyze vast amounts of text and numerical data in seconds, many say the technology has the potential to significantly reduce costs across the industry, but there are fears the disruption could lead to job losses.
“People don't understand that it's there as a productivity tool,” said Nasir Zubairi, CEO of the Luxembourg House of Financial Technology, a fintech advocacy group. “They still really believe it's going to take away their jobs.”
He added: “Traditional banks are fundamentally analogue by design, and converting analogue to digital has always been a challenge.”
Speaking at the Financial Times' TNW technology conference this month, Zubairi gave the example of checking for money laundering: Financial institutions typically hire employees to sift through spreadsheets for unusual activity.
He demonstrated to one agency how he could improve this using a customised AI model, but it was rejected, as he estimated that his method could “instantly save 450,000 euros per year in salaries”.
“People don't want to fire people,” he added. “They want to protect their jobs. And if they have to fire people on their team who perform those functions, they may be under threat as executives, or their power may be eroded in some way.”
Central banks have recently been called on to “level up” with AI, according to the Bank for International Settlements, which says the technology has the potential to bring productivity gains but also carries risks such as providing inaccurate information and becoming victim to hacking.
A common problem with large-scale language models, the underlying technology for most generative AI products, is that they tend to “hallucinate” – state inaccurate information as fact – and they are known to generate information based on the data they were trained on, raising concerns about confidentiality and security of information.
“It's not necessarily going to be rejected. [AI]”But I am hesitant,” said Wincy Wong, head of digital at NatWest, calling for an assessment of the risks, ethics and vulnerabilities of the technology before it is deployed. “At the end of the day, we are one of the major banks and many of our customers store their data and financial information securely in our bank accounts. We need to respect that.”
Customer service is one of the areas most likely to be disrupted by AI tools that can converse and answer questions like a human. For more than a decade, digital banks have used machine learning to screen online questions, often directing customers to a live customer service agent.
But bots equipped with LLM can understand a much wider range of queries, regardless of how they are phrased, and can make decisions such as ordering a bank card, eliminating the need for human intervention.
“I really think the majority of customer service jobs will be gone over the next 12 months to five years,” Monzo's Blomfield said.
Many banks and fintechs, including Klarna and NatWest, are already using AI chatbots for customer service. NatWest's Wong says that their service, AI Cora, has made great strides in generative AI, fielding more than 11 million chats in a year, more than half of which did not require human intervention. In 2017, the service fielded 1,000 chats a month that required intervention.
Swedish fintech company Klarna said its AI assistant can do the work of 700 customer service representatives and resolve inquiries in under two minutes, up from 11 minutes previously, and as a result, the company expects to save $40 million in customer service costs this year.
But Wong says the key to success was training the model to handle nuances, such as understanding that a change of address can have emotional implications, such as a death in the family.
“Understanding the psychology behind it is really important, and if you don't get it right, quite frankly, you could end up upsetting a lot of customers,” she added.
Banks also had to carefully deploy new technologies while adhering to strict industry compliance rules and navigating an uncharted regulatory environment.
In a landmark ruling in 2022, a Dutch court ruled in favor of neobank Bunq, which sued the Dutch Central Bank for banning it from using AI to check for money laundering.
Regulators lifted restrictions on the bank last month following increased scrutiny by German fintech company N26. The bank had restricted new customer sign-ups for years due to weak anti-money laundering practices and faced multimillion-euro fines for repeatedly filing late reports of suspicious transactions.
Karina Kozore, chief risk officer at N26, said the company worked closely with regulators to build an AI model that assesses whether new customers are criminals, reducing crime cases on the platform by 90 percent.
“If we don't adopt AI in our industry, we won't be around in a few years,” she added. “We need to show the benefits we can get from using AI and how we can help you stay compliant.”
