Four ways how AI can benefit the financial services industry

02 April 2020 8 min. read

The advent of artificial intelligence (AI) is providing financial services institutions with major opportunities for improving their client offerings, operations and compliance. Ben Nadel, Director at independently-minded management consultancy Woodhurst, explores four key areas where AI can ramp up effectiveness and efficiency while lowering costs.

Improving customer experience

The rise of challenger banks has demonstrated how technology can drastically improve customer experience. Thirteen million people in the UK use challenger banks today, and that number is set to triple in the coming year. In order to remain competitive, traditional financial service providers are turning to emerging technology, including AI, to improve their customer experience.

Personalised advice
In the UK, Metro Bank, one of the initial challengers, uses predictive analytics to help customers manage their finances. Metro’s in-app insights service monitors transaction data in real-time, and uses this to provide personalised advice to their customers. This type of personal financial management (PFM) capability is the foundation that allows banks to go beyond simple insights and advice, towards the concept of a financial coach in your pocket.

How AI can benefit financial services institutions

This is becoming an AI-enabled reality today. Fintechs such as Plum and Cleo have coupled a chatbot interface with sophisticated AI technologies to create a digital relationship-driven experience with their users, rather than the transactional-driven one of traditional banks. As competition increases and retail banking moves towards digital interfaces, the battle to own the relationship becomes much more important. The AI financial coach becomes the portal through which all other financial products are used.

Improving customer support
Nordea have developed Nova2 to enhance their customer support. Their chatbot uses natural language processing to interpret customer queries and give relevant responses. Customer wait times have decreased dramatically and arguably the responses will be better formed and more accurate. Nordea, and other firms that have introduced similar solutions, will be able to redistribute the people previously responsible for chat towards support issues of a more complex or technical nature.

Onboarding new customers
The traditional experience of opening a new banking product can be drawn out and requires significant effort on the part of the customer. AI is helping to dramatically improve this by streamlining the Know Your Customer (KYC) process. Using image recognition, companies like Onfido can validate a customer’s identity in seconds using a combination of a selfie and a photo of an identification document. This simple integration of an existing technology will reduce drop off rates during the product opening journey, speed up the end to end process for the customer and improve overall customer engagement. 

Fighting financial crime

Financial crime is a major threat to the UK. The financial services industry has invested heavily in systems and innovations that last year prevented £1.6 billion being lost to fraud. Despite this, criminals stole £1.2 billion in 2018. AI provides meaningful insights from data that helps to uncover risks and combat financial crime. This enables human experts to focus on high-probability cases rather than processing thousands of false alerts each day. 

Real-time fraud identification
AI can identify complex fraud patterns and accurately reduce the number of false positives by leveraging large amounts of data. For example, Citi have developed a payment outlier detection service that identifies payment anomalies before payments are processed. Powered by machine learning, this system constantly re-trains against new data in order to improve its detection capability and reduce false-positives. 

Anti-money laundering
Machine learning models can be used to perform analytics and deliver risk scores in real-time. Vocalink and Pay.UK recently launched an anti-money laundering solution for real-time payment systems. Within a few weeks of going live in December 2018, thousands of UK accounts were flagged for further investigation due to suspicious activity — a significant percentage of which were subsequently identified as being involved in money laundering. 

Multiple, large, well-concealed money laundering rings were uncovered. HSBC developed an industry-leading anti-money laundering system which combines customer and counterparty trade information, transactional data and external insights to detect and disrupt financial crime that may have gone under the radar in the past.

Anti-bribery, insider trading, and corruption
AI can identify these forms of financial crime by analysing emails, voice, expense reports and other forms of unstructured data. In both the public and private sectors, AI could be used to analyse data associated with competitive supplier bids, to identify if corruption could have resulted in one supplier being favoured over another.


A lot of time in financial businesses is spent processing paper documents. A manual review of 12,000 annual commercial credit agreements might take 360,000 hours, but the same volume of documents can be reviewed in just seconds by using optical character recognition and text analysis. Laborious processes can be automated by putting AI in the right places:

  • Reconciliation of multiple non digital data sources
  • Data input and reporting across different, often unconnected systems
  • Customer account processing including amending direct debits, or opening and closing accounts

These techniques form part of a larger workflow that reduces overheads by putting tools, processes and applications in a single location. Business process automation streamlines simple, time consuming and repetitive tasks. The result is that people are freed up to work on higher value tasks that involve complex decision making. As AI is applied in more areas of financial business, existing employees can move to those tasks for which humans are uniquely equipped.


Regulatory compliance in the financial services sector is a complex and challenging area. Risk comes in many other forms, particularly financial fraud, money laundering and other financial crimes, and when tackling these sources of risk it’s critical to bring together a diverse range of data sources in order to look for patterns and anomalies. 

While regulation can act as a barrier to the adoption of emerging technology, by the same token AI can be beneficial in making compliance simpler and safer:

  • Automating repetitive tasks can reduce the scope for human error that would otherwise expose the business to increased risk
  • Intelligence-driven compliance assessment can continuously assess compliance by using vulnerability scans, transaction logs, and customer service records
  • Entity resolution is used to connect disparate data sources to uncover risk, possible fraud, and to check regulatory compliance. By breaking down data silos through intelligent linking we gain a single view of customer activity 

Allowing AI to thrive

For artificial intelligence to thrive, it needs a supportive organisational environment to be implemented successfully. If an organisation wants to take advantage of AI, it must review its entire operating model to ensure it is set up for success. Elements of the right behaviours, policies, tools and practices will already be in place, but these should be shaped further to create a conducive environment for AI implementations.

For more information, download Woodhurst's 'Navigating AI in Financial Services' white paper.