How GenAI Sparks Growth and Innovation in Leading Banks

Best Selling Author of Bank 4.0 and Augmented, Host of The Futurists Podcast, Brett King, and Head of Digital Finance for Huawei, Jason Cao resume their teatime chat to bring you more insight on generative AI and the role of regulatory bodies in future innovation.

Brett King: Jason, we’ve talked a lot about GenAI, it being such a hot topic. We’re seeing it being paired with LLMs to create chatbots. Platforms are being used to generate creative content. But we’ve also seen such platforms suffer from hallucinations, which is not something we can afford when we talk about GenAI in finance.

Such platforms need to be trained specifically on financial transaction models, but aside from a few small projects offshore, and some acquisitions, we don't have any stable models for large LLMs for transactional generative finance.

Can you tell us about any projects or work you’re doing in respect to LLMs? And for generative finance, how are you getting the data to build such an LLM?

Jason Cao: GenAI is certainly a disruptive technology. It’s going to bring huge change to society. I think that right now, we’re still in the early stages of the so-called “intelligent world”. But that doesn’t mean we have time to sit around, or to wait and see. We need to keep exploring. The leading financial institutions are already doing this, and are not just doing digital but doing AI. They’ve even sniffed out areas of potential value. For example, one banking giant in China has already implemented more than 100 business cases with GenAI. There’re a few interesting observations we’ve made.

First, we can see they’ve already moved from an experimental stage to implementing it on a wider scale. For instance, where they’ve used GenAI in branch automation, more than 10,000 branches are seeing the benefits already.

Second, business cases are already moving from efficiency-based to value-added. If you’ve established a level of efficiency, you can dramatically reduce how long services take to carry out, and improve their quality overall.

Third, we can see they are moving from singular functions to multiple functions in the form of AI agents. These can perform all kinds of tasks via orchestration and so on, which is of great help to bankers.

Brett King: It would seem to me that banking giants in China with hundreds of millions of customers have got a lot of data to train LLMs, but banks elsewhere do not have that kind of scale. From this data perspective, it seems to me that Chinese banks have an advantage. Is there any effort that you see, either here in China or elsewhere, where banks are saying we should pull our data together to make better generative finance engines? Or is it just bank by bank right now?

Jason Cao: I would say it’s bank by bank. Ideally, if you think about open banking, it would be much easier to compile and have access, but that’s not the case.

However, the real point is data quality. Sure, the more the better, but that by itself is not enough. I think what Chinese banks are doing very well is data governance. This has a huge impact on the model.

Brett King: From a regulatory perspective, do you think the regulators in China will also have to inject AI into their operations and infrastructure? We haven’t had clear guidance on it, but we do have very mature AI regulation already in place.

Jason Cao: Yes, definitely. I think they’re working on that, but AI develops so fast. You cannot have complete and clear guidance for now and for tomorrow. It’s a balancing act across so many parties.

Brett King: You keep talking about speed, but that’s not something particularly associated with regulators. They’re definitely going to have to move faster to keep up with this rate of change. Do you see it as the regulator going from setting policies and creating rules to being a part of the infrastructure itself?

Jason Cao: Yes, this is something I see regulators trying to do not only in China, but worldwide. So, we have to understand what regulatory bodies require technologically as well if we’re going to cater best to the whole industry.

Brett King: All right, let's finish off the conversation with this. In the next 12 months, for your practice at Huawei on the digital finance side, what are some of the things that are exciting for you that you are now wanting to take to the rest of the world?

Jason Cao: There are two things. The first thing that always excites and motivates us is exploring for the future. We need to keep pushing boundaries. The second thing is we’re keen to help our customers tackle technological debt. If you don’t get ahead of that, it will sweep the rug from under you.

There’s a lot to look forward to, of course. It’s like what our rotating chairman said about autonomous cars. He said we will try and provide the best solution that brings the convenience of self-driving cars to everyone. The same goes for finance. We want to elevate the customer experience.

Brett King: Jason, for those interested in what’s happening here in China and with Huawei, could you tell us where they can find out more about yourself and Huawei?

Jason Cao: The easiest way you can keep up is by looking on our website. I just hope that you get the chance to come and visit China and see what it’s like for yourself.

Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy, position, products, and technologies of Huawei Technologies Co., Ltd. If you need to learn more about the products and technologies of Huawei Technologies Co., Ltd., please visit our website at e.huawei.com or contact us.