UBO too slow? AI shows way to stop crime and keep customers happy

In the fight against financial crime, insurers face an uphill battle to ensure their ‘ultimate beneficial owner’ (UBO) checks are effective, exhaustive and efficient. No wonder more firms are turning to AI to tackle the growing volume and complexity, says Joanne Butler at Shift Technology

Understanding who is the ‘ultimate beneficial owner’ (UBO) of an entity or organisation has become difficult and time-consuming for insurers. Complex company structures, inconsistencies in definition across jurisdictions, disappearing UBO registers and growing numbers of new entities are just a few of the challenges.

With over 800,000 new entities added to the UK register alone in the last financial year, the number of necessary checks keeps increasing – particularly for offshore entities which are often exploited by financial criminals to facilitate money laundering.

Insurers are left with a long list of false positives, time-consuming checks and often fewer staff year on year to bridge the compliance gap.

Can insurers stay compliant without compromising the customer journey?

Compliance checks are seen to slow down the buyer’s journey, which is exactly the opposite of what most insurers are trying to achieve, so how can a balance be achieved?

In an ideal world, new insureds and prospective customers would automatically be screened, embellishing the data they have given with internal and external data to build fuller entity profiles. Using policy and claims data, plus external sources including public registers such as Companies House, , the complete information could then be screened intelligently using AI, removing false positives and identifying the UBO with greater accuracy. A greater percentage of quotes could then be fulfilled instantly and “without touch”.

What about personalisation and straight-through processing (STP)?

If there are certain lines of business and/or groups of customers with different expectations of the buying process, segmenting them and giving them different journeys is much easier with the fuller entity profiles built using data and AI. That way customers that are likely to need more information, or more complex lines of business can follow tailored buying paths.

AI can be trained to perform the core checks of an experienced underwriter or compliance official with a fraction of the effort, in real time and without interrupting the buying process.

Joanne Butler, Shift Technology

How does AI help verify the UBO?

AI, machine learning and other advanced technologies can be deployed to complement insurers’ teams. AI can be trained to perform the core checks of an experienced underwriter or compliance official with a fraction of the effort, in real time and without interrupting the buying process.

Sophisticated matching techniques and scoring schemes cut out false positives and score entities based on multiple attributes. Getting a clear prioritised list of items saves significant time on compliance checks and maximises the likelihood of spotting potential financial crime or sanctions evasion.

When budgets are cut and new business volumes grow, the AI seamlessly processes additional cases without needing to add and train more team members. The team are also then able to focus on the cases that are most likely to need their expertise. High-value, complex and ethical cases get the time and human attention they need.

The rules and legislation are constantly changing – how does AI help that?

Rather than constantly add to a complex set of rules in a report-based system which becomes unwieldy and costly to maintain, AI can be trained on many complex scenarios in parallel by simply adding new scenarios as and when they are needed or tweaking existing ones. AI can intelligently apply these scenarios too rather than applying a complex decision tree to all cases, which often results in false positives and missed risks.

Many compliance teams use batch UBO verification. Is it possible to move to real time?

Receiving compliance cases in real time, where the team needs to investigate immediately to move the quote forward, may require a significant process change to be tackled over time. Systems using AI are designed to operate instantly but can be configured to deliver their outputs at more convenient intervals if needed, such as overnight or weekly batch, giving time to reorganise teams and processes.

AI feels like a black box. How is it best explained to the auditor?

While a lot of processing and work done by AI is in the background, the approach, input variables, conclusions and context are all available as outputs. As such, human review and verification can be built in throughout the processing. These rich outputs are produced quickly and consistently and are available for everyone screened, unlike spot checks and sample tests conducted and written up manually which also take significant time to collate for audit.

What’s the first step?

Amid the drive for business growth, superior customer service and increased need for efficiency, achieving robust, cost-effective compliance seems impossible. Leveraging the latest technology can achieve full coverage and real-time screening, while reducing false positives, manual checks and audit preparation, all within existing resourcing. Compliance AI has become a must-have to gain competitive advantage.

Sponsored