Artificial intelligence (AI) technology is transforming the nature of risk and compliance-related roles in the finance sector, effectively reducing the mundane and repetitive processes and giving way to more strategic and creative tasks, according to technology experts.
The growth of the technology is being tempered, however, due to widespread concerns about the impact of tools such as generative AI on high-skilled jobs.
At the recent TechFuture webinar, hosted by Thomson Reuters, speakers described a hesitation among industry professionals to use the technology based on a fear that sophisticated machines will eventually replace jobs. Ivan Perić, senior AI and data analytics specialist at Synechron Technologies, said this was a short-sighted way to look at the challenges and opportunities in the financal services sector.
“The really important thing to remember is that might mean some works are moved to possibly less skilled employment segments with the support of AI, but it also means that on the other end, somebody needs to be the expert that’s training, managing, monitoring, and taking care of the AI products,” Ben Hachey, associate professor and applied AI advisor at Sydney University and Ergo AI, said.
He added that aside from the significant impact on productivity, improved customer feedback, reduced requests for managerial intervention, and better employee retention, AI models may be opening new employment segments focused on ensuring the safe use of this technology.
AI-assisted data validation
Through one of Perić’s recently concluded projects for a bank client, he described how a generative AI model can analyse vast and unstructured data from multiple sources so that it can generate pitch decks for the client’s junior bankers, complete with information such as key people, competitors, locations, customers, suppliers, and all financial information graphs comparing revenues and cash flows.
As the firm’s major request, Perić said they added a backtracking feature to validate those pieces of information, effectively managing the data hallucinations.
“In under a minute, [the AI-generated pitch deck] completes 80% of the work that junior bankers were doing, setting the first step to build something more meaningful in terms of business,” Perić added.
Improved risk assessment
The advanced analysis generated by AI is providing risk experts with deeper insights into potential risks, enabling organisations to make more informed decisions, Perić said.
“One other use case we are tackling right now is supply chains. By finding out which company supplies a given product or services to which company, and monitoring data sources, relevant reports, and news articles over time, a financial company can accelerate identifying trigger points in supply chain.”
“The findings then form the basis to offer some financial services to the client to avoid possible disruption to the business,” said Perić.
He added that “as a human being, it’s basically impossible to read 500 news articles in one day,” suggesting that some patterns and trends generated by AI models may not be immediately evident through traditional risk assessment methods.
Elevating the human factor
Panel moderator Axel Threlfall, editor-at-large at Thomson Reuters, asked the experts whether the importance of human work is being relegated to the corner as generative AI is revolutionising compliance and risk-related jobs, rooted on the premise that “humans, not machines, drive organisational change.”
According to Kate Ingwersen, chief executive – technology, at Challenger, an investment management company in Australia, it’s crucial to support people by providing them with the tools on how to use the technology and unlock its opportunities.
“A real focus for us is, how do we use our people to drive that innovation? How do we use our people to come up with those use cases, those unique propositions because they know our clients better than we do?”
Perić reiterated confidence that AI will not replace jobs in the near future, as human feedback will remain necessary for reliability and “explainability” purposes. He said that risk professionals will always be important in ensuring that AI use aligns with data protection laws and regulatory guidelines.
“These models are put everywhere, and I’ve read an article yesterday that GPT4 performed worse right now than in March. Because of the constant data input, it probably wasn’t cleansed enough. The human element will be needed for all support, such as on privacy concerns and in training custom models inside the organisation,” Perić said.
Raising people’s awareness and educating them on how to use the technology is extremely important for the panellists, partly for Perić, to remove any hesitation on using the tool.
According to Sanjna Parasrampuria, partner (data & analytics expert) at McKinsey & Company in Singapore, generative AI will continue to grow its impact in and out of the financial sector, so however everyone feels about it, there’s only the option to strategise around it as it evolves.
“Unless there are some extreme scenarios from a regulatory shutdown perspective, which you never know, we just got to accept it. What we do with that capability, whether we use it to augment, accelerate, or displace ourselves, is a question we, as a race, will answer as time evolves. It’s all about how you put it to use,” she said.
Parasrampuria’s sentiment is shared by Ingwersen, adding that organisations that will succeed with AI are the ones that usher their people along the journey with them, by giving them the skills and capability to unlock the technology’s value.
“The anxiety of going to be out of a job tomorrow is not there, but a commitment from the organisation to train and skill their people for the future,” Ingwersen said.
Related reading: The Power of AI in Financial Services (thomsonreuters.com.au), AI and cyber-enabled financial crime top risk (thomsonreuters.com.au)
This article first appeared on Thomson Reuters Regulatory Intelligence and features on Business Insight with permission.
