The use of AI in finance was a hot topic at this year’s Sibos conference as banks discussed how AI could change the way financial services are delivered and consumed and, more importantly, how its data is managed.
Using large language models (LLMs) to improve efficiency, improve customer service, and improve decision-making has been part of the conversation since OpenAI launched ChatGPT in November 2022.
While LLMs represent significant advances in AI capabilities, particularly in how machines understand and interact with human language, banks are adopting a cautious response due to concerns about regulatory compliance, data privacy and security, model accuracy and reliability, bias and justice. .
“Banks have always been awash with data, but prioritization is lacking and tags are incomplete or inconsistent,” said Andy Schmidt, vice president and global industry leader for banking, CGI. “To be able to simply train a large language model to search data, you need to have enough confidence in the data and its usability.”
“I think the important part that people need to understand first is to really understand the data. Set up data management and ensure that the data is of decent quality. Being able to eliminate duplication and then figure out where you need to enrich it,” he says.
Standard Chartered provides artificial intelligence solutions and Margaret Harwood Jones, global head of finance and securities, says the bank is working hard to address data management challenges. “You get so many instruction requests in a very unstructured format, so we use AI to convert them into structured data formats that can then be processed efficiently.”
At the Women in Tech Sibos event hosted by EY, panelists discussed how the only way to avoid bias in AI is to teach graduate students to represent everyone from the start, not just white men, and the only way to do this is to have a more diverse workforce .
IBM believes that organizations need to proactively identify and mitigate risks; monitoring fairness, bias and bias. Updates to Granite Guardian 3.0, Granite Telecommunications’ widely used network monitoring tool in North America, allow developers to implement security measures by validating user requests and LLM responses. This includes checking for things like social bias, hate, toxicity, profanity, violence and jailbreaking in the top 10 LLMs.
Because of the potential risks and ethical implications, banks need to take the responsible use of AI seriously, which includes being rigorous about their data.