By Elizabeth Howcroft
AMSTERDAM (Reuters) -The boom in artificial intelligence will increase banks' dependence on big U.S. tech firms, creating new risks for the industry, European banking executives said.
Excitement around using artificial intelligence (AI) in financial services - widely used already for detecting fraud and money-laundering - has soared since the launch of OpenAI's viral chatbot ChatGPT in late 2022 as banks examine ways to deploy generative AI.
But at a gathering of fintech executives in Amsterdam this week, some expressed concerns that the amount of computing power needed to develop AI capabilities would make banks rely even more on small number of tech providers.
ING's chief analytics officer, Bahadir Yilmaz, who is in charge of the Dutch bank's AI work, told Reuters he expected to rely on Big Tech companies "more and more going forward", for infrastructure and machinery.
"You will always need them because sometimes the machine power that is needed for these technologies is huge. It's also not really feasible for a bank to build this tech," he said.
Banks' dependency on a small number of tech companies was "one of the biggest risks", ING's Yilmaz said, emphasising that European banks in particular needed to ensure they could move between different tech providers and avoid what he called "vendor lock-in".
Britain last year proposed rules to regulate financial firms' heavy reliance on external technology companies, such as Microsoft (NASDAQ:MSFT), Google (NASDAQ:GOOGL), IBM (NYSE:IBM) and Amazon (NASDAQ:AMZN). Regulators are worried that problems at a single cloud computing company could potentially bring down services across many financial institutions.
"AI requires huge amounts of compute and really the only way that you're going to be able to access that compute (computing power) sensibly is from Big Tech," Joanne Hannaford, who leads technology strategy at Deutsche Bank's corporate bank, told an audience at the Money20/20 conference earlier this week.
Hannaford said that there are requirements for the bank to notify regulators when they move data into the cloud, which could become much more complicated as the use of cloud computing increases.
Banks also need to communicate to regulators the risk of not leveraging cloud computing’s power, which would be an opportunity cost, she added.
AI was top of the agenda at the Amsterdam conference.
The CEO of French AI startup Mistral AI, seen as France's answer to OpenAI, told attendees there were "synergies" between its GenAI products and financial services.
"We see a lot of opportunities in creating analysis and monitoring information ... which is really something that bankers like to do," Arthur Mensch said.
ING is testing an AI chatbot currently used for 2.5% of incoming customer service chats. Asked how long it would be until the chatbot could handle half or more of customer service conversations, Yilmaz said within a year.
In its first statement on AI, the European Union's securities watchdog said last week that banks and investment firms cannot shirk boardroom responsibility and have a legal obligation to protect customers when using AI. It warned that the technology is likely to have significant impact on retail investor protection.