Banking is about to change
It is difficult to move without talk of AI and how it’s going to shape the future.
While AI has been developed since the 1950s, it is only in the last 3 years that it has risen in public consciousness, with eye-watering investments in technology companies and the emergence of “human like” chatbots such as ChatGPT. The challenge is separating the reality from the hype and understanding how AI will fundamentally change the banking sector in the next 10 years.
AI brings with it huge opportunities but also brings into question long-held business practices – with a gulf opening up between disruptive “challenger” banks with massively reduced operating costs and traditional banks with reliance on buildings and staff.
AI makes technology accessible
For years, jobs that humans found easy, software found very difficult (counting people in an image for example). The new generation of so-called Generative AI changes this.
AI “Agents” can answer customer queries, check legal documents or create images from a written description. They are able to augment or replace a wide range of routine tasks, improving customer service and reduce costs.
These changes are as fundamental to banking as the Internet or mobile technology. Like the Internet, it will shake up the status quo and create both winners and losers. Deciding how, and how fast, to adopt AI is one of the main challenges for Boards and senior management of banks in the next 5 years.
Case History: AI in action
Generative AI has been recently employed by Citigroup to evaluate the effects of new US capital regulations. The bank’s risk and compliance team used generative AI to sift through and summarize 1,089 pages of new capital rules released by the federal regulators.
Klarna’s customer assistant handles two thirds of all customer chats (over 2.3 million enquiries). Available in 23 markets and 35 languages, it reduces average resolution time from 11 minutes to 2 minutes.
Understanding AI – Video Resources
- Introduction to AI – TEDx
- Putting AI to work in Finance – IBM
- Generative AI – Google
- Uses of AI in Banking – Vision Economy
- AI Agents – Andrew Ng
- How AI will Disrupt Banking – Forbes
- Technical overview of AI Agents – AI Software Dev
01
Predictive vs Generative AI
For years, the role of AI was prediction. Teams of data analysts use technology to mine data in order to obtain insights, where the AI is trained on massive datasets to “learn” the patterns in the data. This is the basis of self-driving cars, Google Translate and many stock trading platforms. Predictive AI is powerful but complex and expensive to implement. It requires specialist teams, huge quantities of data and can take many years to develop.
Generative AI is different. Here the AI has already been trained on human language. Its role is to determine what comes next in a set of information. It uses this to answer questions such as “what is the capital of Kenya” but also much more complicated questions like “write me a letter of complaint about the phone I just purchased” or “does this contract have a termination clause?”. Generative AI can be used directly and does not require training.
02
The rise of Agents
Just as generative AI can answer a question, it can also create a plan and access tools such as a calculator or a web browser. In this way, AI Agents can plan, obtain information and execute tasks, making judgements at each step, much as a human would, even when the information it receives is poorly defined. For example, AI agents can read an email, determine the sentiment and topic and either respond with information or route the email to the relevant department. Agents have the potential to transform the operations of the banking sector, specifically automating tedious but poorly defined tasks and freeing staff to undertake more rewarding activities.
03
Ethics, Governance, and Compliance
While AI may appear to be just another digital technology, its use is very different to conventional software projects. AI has additional regulatory requirements and may exhibit unreliable behaviours if not correctly managed. Unlike software, simple testing is not enough – AI has to be designed for robustness and have safeguards built in from the outset. This is the new horizon of AI – balancing functionality with assurance and resilience.
At its core, AI sits at the junction between technology and business needs. AI technology is evolving at a rapid pace and has to link to the existing organisation. This poses a challenge, balancing the needs of new development with maintaining existing services.
Many banks are using AI without knowing it, from staff using ChatGPT to automated CV scanning in HR. While these bring benefits they also expose the bank to potential risks regarding leakage of confidential customer data or regulatory sanction.
AI is not just a technology issue but touches almost every aspect of the bank – from ethics, to cybersecurity to PR. Boards are increasingly assigning specific responsibility for AI within the business, such as the creation of a Chief AI Officer role and formalised processes for handling AI in a consistent way across the business. In addition, widespread staff training is needed to make members aware of the considerations concerning AI and to provide a clear route to reporting any issues that arise.
Getting started with a AI
While AI may appear daunting, it is easy to get started with a simple clear approach:
- AI audit – What AI does the bank use today and how are they managed
- AI Governance – Put in place a simple but effective AI governance process
- AI Opportunity Discovery – Map the places where AI can have most impact
Getting started with AI is about choosing the right problem. To begin with, choose something straightforward, that adds value at low risk. For example, provide a natural language tool that helps loan officers process customer applications more quickly, reducing customer wait time and making staff more productive. Build your expertise with AI and start the journey to AI Digital Transformation that will change the face of banking.