Generative AI is now becoming a reality within banking.
McKinsey found more than 50% of banks in US and Europe are adopting “more centralised” generative AI in their organisation.
The debate, or rather big neon pink elephant in the room, is whether generative AI advances will improve efficiency or replace jobs. This is emphasised by a
recent study by Citi, which showed that up to 54% of jobs in banking have a high potential of automation, higher than that of other industries.
In May 2024, Klarna
announced that 90% of its employees were already using generative AI in their jobs, creating greater internal efficiency. This comes after
the company announced its OpenAI-powered chatbot is doing the equivalent work of 700 full-time employees.
I spoke with Martin Elwin, Klarna’s Head of AI and Shanker Ramamurthy, Global Managing Partner Banking & Financial Markets, IBM Consulting about generative AI’s impact on the workforce.
Shifting workforce
“You are either developing generative AI, using generative AI or generative AI is going to use you,” Ramamurthy told me during a conversation at Money 20/20.
A recent
IBM survey of over 297 CEOs in the business and financial markets from over 30 countries showed that the jobs in the financial sector are changing. 50% of the financial CEOs surveyed said they are hiring for roles that did not exist last year due to generative
AI.
Ramamurthy commented: “From a financial services standpoint, it is very clear from the study that CEOs are racing to adopt generative AI. This is despite the challenges that they have in terms of skilling and scaling their workforce.”
Outside of the challenges in ensuring that AI they use is safe from a regulatory perspective, companies are facing employee challenges. 53% of CEOs surveyed said they are struggling to fill key technology roles.
Filling the roles is just one thing, tackling the cultural and skills change AI brings is another. Ramamurthy told me: “About two thirds of bank CEOs were talking about how their employees are concerned that technology is changing faster than their ability
to actually adapt.
“On the one hand, employees are concerned, on the other hand, CEOs want more adoption. So there’s a whole change management challenge, which is industry wide.”
See below for some of the other data the IBM study highlighted.
Source:
IBM Annual CEO 2024
Banks perceive that if they do not use AI they will ‘lose’, with 57% of respondents believing that competitive advantage will depend on who has the most advanced generative AI.
Efficiency or threat to jobs
Klarna made headlines by announcing their new chatbot could do the work of 700 fulltime humans. However, Elwin pointed out to me that they were outsourcing their customer service, meaning they aren’t lowering their internal numbers.
Klarna does have a hiring pause at the moment because they think AI can fill in the gaps in a lot of jobs. Elwin elaborated on this: “It’s not so much that AI directly replaces the need for people. We are pausing hiring for a bit to see the effect of us
using AI across the organisation because it will definitely have an impact on the needs of different kinds of skills.”
Elwin described one example of where they have implemented AI in their financial department to improve efficiency. In this use case, the departments used AI to help them with the quality control assurance for the written parts of internal reports.
He said: “It just helps shorten review cycles, it helps get to the quality level. We want to be at quicker it helps them more quickly deliver on those reports.”
With the majority of their staff now using generative AI in some way, Klarna are claiming to see increases in efficiency, and in some places lowering costs. They have stated they have managed to
cut marketing costs by 25% while running more campaigns.
Klarna seem to be looking at AI more as a tool, which may cause shifts in how jobs are done, but ultimately make them more efficient. As Elwin stated: “Just because we had one plan previously, as we apply AI, it might shift a little bit. How we best use
those resources or where we actually benefit the most from hiring people.”
Despite these efficiency improvements, data presented by Citi from an
earlier Accenture report, is showing that finance jobs may be some of the most threatened by the progression of AI and generative AI.
Source:
Accenture, ‘A new era of generative AI for everyone’
While at the moment generative AI may be creating better efficiencies, with a forward looking lens there should be some cause for concern in the job security of certain banking positions.
Efficiency over trust
One statistic within the IBM study which did cause some concern, is that 66% of business and financial CEOs stated that the potential productivity gains from automation are so great that they would accept significant risks to stay competitive, with 67% saying
they would risk more than their competitor to maintain competitive edge
Given the current emphasis on generative AI regulation and safety, this is deeply worrying. IBM themselves have an emphasis on all of their AI being explainable.
Ramamurthy said: “You need to be able to prove to regulators on the one hand, that any decisions made by generative AI are explainable are within the bounds of the regulations, which is a non-trivial challenge. On the other hand, you also want to be, as
an enterprise, confident that the tools and the data using from a large language model is not going to put you in a situation where you could be sued for using somebody else’s data.”
The singularity? Not quite yet
At Money 2020 Europe 2024, there was a stage called the Singularity stage. I was thrown by this, because to my mind, singularity is not often a good thing. It evokes images of Terminator, I Robot or Black Mirror, dystopian futures controlled by robots and
programmes. Yet for some, this is a utopian concept, no work or reduced work in a more intelligent world. Much like that fiction this is all quite speculative.
Bringing down from those lofty concepts to the reality of our situation, at the moment generative AI is a tool. It can be used to help people to their jobs more efficiently and effectively. Examples like summarising documents, writing the base of a contract,
or giving customers simple responses are all extremely useful in our current world.
However, financial institutions are at their heart are profit driven enterprises. Where a job can be effectively cut, it’s more than likely, over time, that it will be. I am not immune to this. I work in media, another sector where a lot of the work I do
can be automated, I already use it for many of my tasks.
For those worried about their positions, you may want to think about training yourself on AI. Adapting to the circumstances and companies should be upskilling their employees as well. The running of these companies is nowhere near that utopian ideal, and
I personally don’t think the technology is there yet either.