Pairing AI and human intelligence in the fight against fraud

April 11, 2024

Share on LinkedIn Share on Facebook Share on X

By Ian Carman, Director of Investigation Services

Artificial technology (AI) and other technological advancements have been adopted in the claims industry to some extent during recent years — and we can safely expect that trend to continue. As new discoveries are made regarding how AI can be leveraged operationally in various lines of business, we must be proactive in assessing its benefits (and risks). 

In a landscape of rampant fraud, humans and machines can forge a symbiotic relationship, building upon each other’s unique capabilities to deliver success for insurers and keep the cost of premiums stable for genuine customers.

Clear benefits

Today, some companies are using AI technology to protect customers from threats — even ones they might not be aware of yet. For one, it protects against the commonplace practice of “ghost” brokers carrying out identity fraud and using customers’ identities for fraudulent purposes. It also identifies false document fraud which could include doctored invoices and manipulated images.

Technology can identify potential fraud far earlier in a claim’s life cycle than the eye of a human investigator. When deployed well, fraud detection technology will fast-track genuine claims settlement, and if it does detect fraud, halt payments from going through. Quicker and more accurate claims processing can keep costs low for insurers and keep premiums down for genuine customers. 

This is playing out in the motor industry already: machine learning (ML) technology is in use that helps assess potential fraud risks — and determines whether the risk is viable — much earlier. As a result, an insurer could fast-track genuine motor claims to settlement and squash fraudulent ones.

Tech-enabled early fraud detection enables investigators’ decision-making and overall responsibilities. As technology weeds out unmistakably fraudulent claims — with less room for error — investigators can conclude and move on to the next case sooner, rather than contribute additional hours resolving the investigation. Fraud detection technology also, notably, saves time and money by pinpointing the exact root of the fraud, so investigators do not have to.

Weeding out fraudsters more efficiently aids in protecting a brand’s reputation, too, both within the industry and from a customer’s perspective.  

Machines versus humans?

We are all familiar with the ubiquitous moral and economical question that looms large over the AI debate: Will machines eventually render humans obsolete? To that end, shouldn’t it be feared, not embraced?

One universal truth helps bolster a valid argument to the contrary: machines cannot feel. Experiencing emotions is unique to humans, and impossible to learn. Complex emotion and reasoning are integral to business: Empathy. Mutual understanding. A sense of duty to do the right thing. Interpersonal communication. Negotiation. Trust. 

Conversely, technology can complete certain processes and tasks, such as the identification of a forged document, quicker than any human brain. Pairing human and machine qualities together can more effectively fight fraud. Tech can excel at identification, while the latter piece involving the legalities, the complexities, the nuances, the interpretation — can stay in the wheelhouse of human claims professionals. For certain aspects of counter-fraud, humans will likely always be involved.

Doubling up

Insurers are constantly screening claims data. Still, data indicating fraud slips through the cracks. When an insurer passes the baton to a business such as Sedgwick, continuous screening can continue, and from a significantly larger pool of data. 

In the commercial property arena, for example, a severely limited pool of data exists to screen against. Less data is disadvantageous for identifying patterns of serial claimants and unwanted fraudulent behaviors in the industry; utilizing a single screening tool hampers the ability to recognize the true volume of fraud frequenting an insurer’s books. 

Potential disadvantages

As technology advances, customers and claimants are increasingly adopting the assumption that technological advancements equal quicker and more efficient processes, and therefore claims should result in faster settlement. Practitioners must keep this in mind: the widespread expectation of quicker resolution — regardless of a case’s intricacies — is now weaved into the foundation of the tech-enabled claims process.

Claims handlers should apply additional diligence with spot checks to meet these rising demands, and intentionally take advantage of software that can aid them.

Learn more > read the flyer to learn more about our fraud claims service and catch up on our latest fraud blogs

Tags: AI, AI technology, Artificial Intelligence, Claims, claims technology, Fraud, fraud claims, fraud strategy, Insurers, Preserving brands, Risk, Technological advances, Technology