Skip to Content

Implementing AI for everyone: Banking’s #1 new year’s resolution

January 28, 2020

According to Barclays, the AI market will likely be worth $89.8 billion by 2025. AI is already improving customer experience, but what is the larger impact of AI on the financial sector, and how can AI be leveraged effectively throughout the back-, middle-, and front-office?

AI isn’t just for customers

Customer service

Customer service is no doubt the area where AI currently has the biggest impact. Chatbots are already able to respond to simple queries from clients and carry out basic tasks such as creating or cancelling a standing order or direct debit and have significantly reduced call center volumes.

This success means banks are now moving from natural language processing (NLP) bots to intelligence-driven AI assistants – leading to faster resolution times and a 30% decrease in customer service costs. Key examples of chatbots in the market right now include Bank of America’s digital assistant Erica and Capital One’s Eno.

Fraud detection

AI also helps prevent and detect fraud by flagging unusual transactions and is capable – if instructed to – of detecting and monitoring unusual behavior. In short, combining supervised and unsupervised machine learning with a more comprehensive, AI-driven fraud detection strategy enables banks to quickly and accurately detect automated and increasingly complex fraud attempts.

For example, Danske Bank recently modernized its fraud detection process and reduced the amount of fraud detected by 1,200 claims a day. By revamping this process, the bank was able to reduce its false positives by 60% – and is expected to reach 80% as the new machine learning model continues to learn. Danske also increased detection of real fraud by 50%, and can now refocus time and resources toward actual cases of fraud and identifying new fraud methods as they appear.

Reduce human intervention

Finally, AI can also reduce human intervention in the payment process and could make onboarding procedures smoother. With AI, NLP processes can read key documents, make sense of them, and give their findings to human decision makers. The result will be a faster, friendlier onboarding process – as banks are able to focus on welcoming new customers and not on the paperwork.

For example, HSBC has invested $2.3 billion on a digital platform and AI capability to reach tech-savvy customers. These AI capabilities also help to handle thousands of invoices and trade documents. In addition, Standard Bank of South Africa uses a WorkFusion AI-backed solution to reduce customer onboarding time from 20 days to just 5 minutes.

All of this means banks will need to change how they do business. According to  the World Retail Banking Report 2019 from Capgemini and Efma, banks need to move away from discrete products and focus on delivering a great overall experience to their customers and employees. Implementing AI in the key areas above will go a long way in achieving this.

AI in the front, the middle, and the back

Banks have now reached a crossroads. Fewer than half (38%) say they have the necessary digital and leadership capabilities required for transformation. But transforming is not impossible. Just remember, it’s not all about storing large amounts of data or just being agile. To ensure a successful AI initiative, certain things should be centralized. This includes excellence hubs for machine vision and NLP, which are critical in supporting AI initiatives and vendor assessment throughout the company.

Critical AI-based governance functions should also be centralized, making it easier for financial institutions to sustain a global competitive advantage and regulatory compliance. Data architects and cybersecurity experts should also be in this group.

Next, only your cross-functional agile teams should be responsible for developing AI-enabled processes, products, and services. AI experts and topic specialists are your core driving force here.

Finally, do not forget anyone managing AI-modified processes and actions. These units are the ones that see the real-world implications of AI in action. As such, they will need to understand any new tools that emerge in the market and how they affect processes and skill requirements. This information should come directly from your global IT team.

What’s driving this disruption?

Enterprises prefer buying AI, not building it. Nearly 50% of companies favor buying AI solutions from third parties, while 33% build their own custom solutions. Meanwhile, 10% of companies will wait for AI to be incorporated into key software products.

But this is where the new kids on the block come in. “If a traditional bank needs to build an innovative solution, 12 to 18 months may be required to handle the project in-house. However, many FinTechs can offer a 60-day turnaround time (here),” according to Vik Ramesh, CEO of California-based startup Fintel Labs.

This is why FinTechs are no longer seen as a threat to established banking firms. They are now a key partner in the race to implement AI, due to their ability to operate at the frontier of financial technology.

In conclusion

AI is critical for building a competitive global advantage in 2020, with many AI banking leaders focusing on customer service, fraud detection, payments, and employee onboarding.

But how can you implement AI for both for your employees and customers? Simple. Just ensure key excellence hubs and governance functions are centralized, the right teams are developing your AI processes, and that your front-line team members get all the support and help they need from your global IT team.

For more information on how to implement AI in your organization, please feel free to contact me, Rishi Vijay


Rishi Vijay, Senior Director and Global Account Executive, Capgemini Financial Services UK:

Rakesh Roshan, Director – Digital and Cloud Practice, Capgemini Financial Services: