Brought to you by:

PBT Group flags importance of “data maturity” in detecting insurance fraud

Data and analytics specialist PBT Group says contributions from Artificial Intelligence (AI) and Machine Learning (ML) can help insurers better detect fraudulent claims.  

The group says that by understanding fraud trends and underwriting data, automated systems can help reduce “burgeoning costs” associated with false claims and decrease premiums.  

Insurance fraud is estimated to cost the industry upwards of $2.2 billion annually, according to the Insurance Council of Australia’s fraud bureau.  

“While AI and ML can automate fraud detection, several factors impact on the accuracy of AI/ML models, such as data quality and variety,” PBT Senior Data Analyst Nitin Mistry said.  

“However, organisations can overcome those challenges by improving processes and incorporating external data.”  

A Federal Government sanctioned review by health economist Dr Pradeep Philip in March found that Medicare held an “insufficient capability” to produce “a proactive, prevention and early detection-focused approach”.  

Mr Mistry says that in his analysis of Dr Philip's research, he detected “a host of data issues that are likely replicated in many insurers”.  

The report recommended implementing an analytical model to help Medicare detect opportunistic fraud promptly.  

Mr Mistry acknowledges that building an automated model “is not an overnight, one-off exercise” but says the results of it can help insurers identify and prevent fraud more efficiently.  

He says businesses would need to organise a set goal, implementation plan and performance indicator for their AI operations and ensure that it is regularly updated.  

“Fraudsters don’t have just one page in their book, so it’s essential to keep up to date with their scams and retrain the model accordingly,” Mr Mistry said.  

He says that, as with most AI operations, “human oversight is essential to ensure the model’s accuracy and reliability”.