Consider that you are in a jungle north of South Africa’s Cape of Good Hope watching an ostrich being stalked by a few tigers. The ostrich, realizing this menacing presence, does not worry as it knows what to do: stick its head in the sand and the tigers will go away. (1)

That is retail banking today, where the ostrich is a retail bank and the tigers are the mobile- only AI powered challenger banks (not to mention Amazon).

According to Business Insider 40% of millennials do not visit physical banks at all. Most actually consider a visit to a branch as ridiculous. The retail banking business model is clearly out of sync with today’s economy, technology and consumer demographics. Already traumatized by accumulated Non-Performing Loans (NPL’s), it is not just an issue of digitization and simply providing a web banking service over mobile wrapped as an Android or iOS app. Although consumer facing digital banking is not something new and includes ATM’s and legacy web banking services, the time for an AI powered next step in banking has arrived, and it is indeed a large step.

What to do?

Essentially, we are now in a different reality where digitization, in conjunction with Artificial Intelligence, needs to be urgently addressed as a transformative implementation vehicle for a new paradigm. This reality is reflected by this LinkedIn Post on the top 10 UK Startups where 60% of the companies mentioned are in the Fintech Challenger space.

So, what to do? Top Banking leaders and/or board members cannot be expected to know AI and where, how and when to leverage it. However, ultimately they decide and the risk of not doing anything is substantially more than the risk of doing something.

Proposed approach

Let’s outline a plausible approach. Firstly, an organizational structural unit needs to be created, reporting directly to the top (CEO and/or Board). Built as an AI Task Force led by a technology AI “visionary”, a.k.a. Chief AI Officer (CAIO), this leader needs to be a doer with real world practical AI experience (not necessarily focused on banking/finance).

The CAIO leader will then create a small team where each team member will address a specific AI technology that will entail web, mobile and IOT implementation and distribution channels (i.e. ‘AI first‘). This organizational structure in itself will reflect the Bank’s new AI Strategy.

So, let’s look at an initial proposed mix of target areas and accordingly each team member’s portfolio in terms of benefits, implementation details and real-world examples.

Team Member 1

Portfolio: Conversational AI

Aim: Build the bank’s fundamental Digital Assistant

Description:
The heart of the matter, the bank’s digital assistant will eventually take center stage. It will be deeply rooted in AI reasoning and natural-language understanding and generation. It needs to be able to handle:

• Sophisticated questions about finance management
• Customer support interactions
• Teach customer basics (like how to open an account, use humor)
• Robotics at physical branches for old-school throwback visits

An example is Kasisto

Team Member 2

Portfolio: AI CRM / Personalization

Aim: Suggest and provide tailored products and services whereby a sale is always closer through the creation of Personas that provide individual and personal customer experiences.

Description: Banks hold huge amounts of customer data in addition to balances, investments and loans. Additionally, Big Data is available in relation to real-time customer behavior from online and offline purchases, website visits, engagements via kiosks, email exchanges and mobile apps. This data is a gold mine where mining is implemented through predictive machine learning based AI algorithms that can identify correlations between customer gender, income, age, purchasing behavior, preferred location of service and mode of interaction.

All the above can be fed into a corporate AI CRM that will function as a central point of reference, from which to additionally:

• Personalize content (focus on micro-elements)
• Retarget customers with new more relevant ads based on their interactions
• Real time event personalization – i.e. what to do if a cart is abandoned
• Use Sentiment Analysis on product customer reviews and retarget accordingly
• Customer segmentation for personalized targeting

Team Member 3

Portfolio: Machine Learning: Non-Performing Loans (NPL) / Risk Management

Aim: Risk status assessment of an account, identity potential NPLs

Description: A non-performing loan (NPL) is a defaulted loan in which the borrower has not made the scheduled payments within a specified period.
AI/ML algorithms can automatically assign an NPL risk score to a customer as part of a general score that reflects overall risk, activity anomalies and high-risk behavior with a view to dynamically predicting potential NPLs.
When risk is identified, preventative action can be taken such as:

• Increasing the frequency of risk assessments
• Adding guarantee collaterals
• Requesting earlier repayments

Overall, NPL prediction can reduce risk, lower reserves and increase profitability, making this team member’s work directory calculable in terms of bottom line profits.

Team Member 4

Portfolio: Biometrics

Aim: AI Powered Authentication / Security

Description: AI Powered biometrics like facial and voice recognition draw upon huge amounts of data to fine-tune authentication. The security implications are huge and self-evident, but this technology can also help with customer service.

As an example, Natwest uses Selfies to open current accounts within minutes. New customers submit a picture of themselves and a photo ID. Through an AI-powered real-time biometric check, the two images are compared in order to authenticate the ID. This drastically increases efficiency and reduces fraudulent applications.

Team Member 5

Portfolio: Banking as a Service

Aim: Create a community of third-party developers and innovators working with the bank’s IT systems

Description: Amazon and Google are probably(!) thinking at some point of becoming banks. One way to hedge around this is to allow third parties (given customer consent) to securely access account data with a view to enhancing the services the bank provides with a view to enabling small accredited niche vendors to imaginatively innovate. This can be done by building a platform through which to enable the development and provision of bank approved web, mobile and IOT apps, much like the iStore (Apple) and the Play Store (Google). Through this digital ecosystem, bank customers will be able to integrate financial services into their offerings and build their own scalable banking products.

Team Member 6

Portfolio: Trading / Wealth Management

Aim: Implement Algorithmic Trading

Description: Implement machine Learning based selection of potential stock trades based on short term stock price predictions updated in real time. AI Powered monitoring of Volume Spikes, Moving Average breakouts, Trend and Support / Resistance breakouts and extreme intraday fluctuations. Trade based predictions in real time.

Conclusion

Technology and especially Artificial Intelligence is shaking up the banking industry. For any retail bank that wants to be around in 2030, there is no choice but to plan and implement a way forward that embraces and maximizes benefits from AI technology.

Takeaway

Fully transforming a bank using AI technology as an implementation vehicle can be done by creating an AI Strategic Task Force that will develop and carry out an AI Strategy that implements specific actions which correlate to the use of the best suited available eAI Techniques.

The aim is to dramatically improve performance and efficiency in particular areas of business activity like customer interaction and support, loans, security and investing.

In all implementation areas the software developed will interact with each of the other areas forming an AI Engine which itself will interact with other legacy banking systems, so forming the backbone of the reborn banking business.