For Finance Firms, It’s Too Late to Not Trust AI

Financial institutions can confidently embrace artificial intelligence (AI) to improve vast numbers of real business use cases, while also satisfying concerns from compliance and governance teams. But it is rapidly becoming too late to simply not trust AI and allow caution to inhibit innovation.

Don’t just take my word for it; let’s look at some examples in the industry from its top leaders. JP Morgan Chase, the world's biggest and most successful bank, has shown the way forward with how it has embraced AI, ensuring full integration from the top down and a head of AI who reports directly to the CEO. JPM also now devotes $2 billion of its $18 billion annual IT budget to AI developments across over 600 use cases, and it expects those to double or triple over the next couple of years.

AI is too important

However, it’s not just the money that’s important in a financial institution; it’s the culture and philosophy that need to change. Some leading banks have taken AI and data out of technology groups and have them reporting directly to the board. They know that, moving forward, there will be no job or business function that won't be affected by AI.

Partner support

Nevertheless, nobody suggests this will be easy or straightforward. It is why at SoftServe we have built up teams of talented AI experts. They have proven experience working with both leading financial institutions and the top providers of AI and associated technologies.

Data access

AI works best when it has a lot of data to work with, and accordingly, to be successful, it needs to access and engage across datasets within an organization. It should not be confined to data that has been set aside for a PoC. One of our first tasks is therefore to work with firms to make sure that data is clean and has the proper foundations and guardrails built to safely maximize the outputs when applied to large language models (LLMs).

Legacy systems

Of course, many banks and insurers might believe they do not have the optimum IT environment to deploy AI, given the state of legacy infrastructure and systems. But this no longer needs to be the barrier it once might have been, as we found out working with many firms to both optimize and upgrade systems in preparation for AI.

We have even developed our own Generative AI code modernization toolkit with multiple supported business cases. One of these is an accelerator for COBOL translation into Python for easier programming and configurations moving forward.

The integration of cloud platforms with on-premises resources to create a diverse, multi-hybrid approach becomes another key foundation for AI to flourish with the use of more sophisticated data analytics. It is why we have developed close working relationships (and often premium partner status) with leading hyperscalers like AWS, Google Cloud, and Microsoft Azure. It means we have done the hard yards, so our clients do not have to reinvent the wheel and can help you determine which approach works best for your organization.

The next step is to decide on what the data/digital strategy is to deliver the outcomes required. SoftServe has the experience to talk with clients at the highest level about how the business can extract maximum value from the organization’s IT assets.

This is not a one-off process. Data strategies need to be revisited at least quarterly, not annually, as most do, or even only every 2-3 years, as many have been comfortable reviewing. By embedding this activity, it will become second nature. Many large banks already go well beyond minimum expectations, for example, by embedding almost continuous stress tests, when regulators only require them to do so once a year.

Competitive edge

The whole purpose of this groundwork is to lay the foundations for a lasting competitive edge, which not only grows revenues but also drives permanent efficiencies. And, while AI models can be trained on an organization’s specific data, it needs to be something more than just ChatGPT if it is to provide a sustainable point of difference.

One area where real advantages are being seen from AI is in payments. There are significant and onerous compliance and KYC obligations on merchants and service providers to meet the standards required by the likes of Visa and Mastercard.

These are not only from the outset when joining a scheme but must be maintained on a weekly or even daily basis as new terms and conditions are acknowledged and implemented. In many firms, these are still done manually as changes to paperwork are entered into systems. But AI can be trained to complete these tasks automatically, saving thousands of work hours, allowing people to be deployed on more valuable contributions.

Conclusion

The key to being successful with AI is getting people involved — everyone. Don’t allow teams to be threatened by AI; make them feel empowered by how AI can help them do their jobs better and deliver superior services for customers.

Once this is underway, a majority of staff will be working with AI and LLMs. It means they can use the bank’s data to deliver superior and personalized services to customers. It means confidence in being compliant with fast-moving regulations. As we often hear from those who have embraced this new approach, “Unless you use technology to do a better job, you’ll be left behind.”

Of course, not every financial institution has the resources of a tier 1 financial institution. It is why at SoftServe we have built the proven, experienced skills to help banks and insurers scale up (and down) when they need specialist AI and other technology expertise.

However, getting the technology right is only half the battle. Business cultures in many organizations will need to change to fully embrace AI and send that message out from the top down. SoftServe works closely with senior leadership groups to smooth these transitions and maximize the benefits. Get in touch for a conversation to discover how it can work for you.

Start a conversation with us