How financial institutions can avoid AI induced hallucinations

By Barley Laing, the UK Managing Director at Melissa

The transformative power of artificial intelligence (AI) promises to, and in many cases already is, driving real-time business success in the financial services sector, from enhancing data-driven insight and productivity, to improving fraud detection and customer experience.

Yet, in the rush to adopt AI many haven’t factored in how having poor quality data on customers could lead to gibberish, or worse, biased and inaccurate outcomes. What we call AI induced ‘hallucinations’, which leads to poor results. 

For example, if someone living in a wealthy area of a city applies for a loan, and the postcode the bank has on their system for them is in an area of the city that is less prosperous, this could impact on the value of the loan offered and the rate of interest calculated by AI. This may encourage the prospect to source a quote elsewhere. 

Decaying data

Data decay is a major factor impacting on the effective implementation of AI. Data decays swiftly with user contact data lacking regular intervention degrading at 25 per cent a year as people move home, die and get divorced. Also, 20 per cent of addresses entered online have errors; these include spelling mistakes, wrong house numbers, and incorrect postcodes.

Inaccurate contact data can be avoided by having verification processes in place at the point of data capture, and when cleaning held data in batch. This typically involves simple and cost-effective changes to the data quality process.

Address autocomplete / lookup for accurate data in real-time

An address autocomplete or lookup service is a valuable piece of technology to use at the customer onboarding stage. It delivers accurate address data in real-time when onboarding new customers by providing a properly formatted, correct address when they start to input theirs. It also reduces the number of keystrokes required, by up to 81 per cent, when typing an address. This speeds up the onboarding process and lessens the probability of the user not completing an application to access a service, for example. This approach to first point of contact verification can be extended to email and phone, so that these valuable contact data channels can also be verified in real-time.

Deduplicate data with an advanced fuzzy matching tool

Data duplication is a significant issue, with duplicate rates of 10 to 30 per cent on customer databases not unusual. This often happens when two departments merge their data and errors in contact data collection occur at different touchpoints. Not only does duplication have the potential to confuse an AI application, but it adds cost in terms of time and money, particularly with printed communications, and it adversely impacts on the sender’s reputation.

Deduplicating data using an advanced fuzzy matching tool is the answer. By merging and purging the most challenging records it’s possible to create a ‘single user record’ and obtain an optimum single customer view (SCV) that AI can make learnings from. Also, organising contact data in this way will maximise efficiency and reduce costs, because multiple outreach efforts will not be made to the same person. A further benefit is that the potential for fraud is decreased because a unified record will be established for each customer.

Data cleansing supports AI tools

Data suppression, or cleansing, using the appropriate technology that highlights people who have moved or are no longer at the address on file, is a critical element of the data cleaning process, and therefore in supporting efforts with AI. As well as removing incorrect addresses, these services can include deceased flagging to halt the distribution of mail and other communications to those who have passed away, which can cause distress to their friends and relatives. By implementing suppression strategies organisations can save money, protect their reputations, avoid fraud and support their AI efforts.

Source a data cleaning SaaS platform

Today, it’s never been easier or more cost-effective to manage data quality in real-time to support AI and wider business efficiencies. It’s possible to source a scalable data cleaning software-as-a-service (SaaS) platform that requires no coding, integration, or training. This technology cleanses and corrects names, addresses, email addresses, and telephone numbers worldwide. It matches records in real-time, ensuring no duplication, and provides data profiling to help source issues for further action. A single, intuitive interface provides the opportunity for data standardisation, validation, and enrichment, resulting in high-quality contact information across multiple databases. It can do this with held data in batch and as new data is being collected, and can also be accessed via cloud API or on-premise, if required.

AI can give your financial institution a competitive edge, but this is dependent on the quality of data fed into the AI models. Poor data leads to AI ‘hallucinations’ with unreliable predictions and therefore bad outcomes. To maximise the success of your AI efforts implement best practice data quality procedures. 


Most Popular