The Ethics and Use of Artificial Intelligence & Machine Learning in Fighting Insurance Fraud  

For decades, the insurance industry has utilized data or historical consumer information for underwriting, pricing, risk detection, and the identification and investigation of insurance fraud. The understanding of consumer risk and potential loss before pricing and underwriting an insurance policy is fundamental to the agreement between the insurance customer and the insurer. To set appropriate parameters for an insurance policy, information is requested, provided, gathered, reviewed, before the terms of a contract are finalized and a policy is issued. As the collection of sensitive consumer information increases and the use of data becomes more sophisticated, there needs to be a conversation to establish standards for how data is utilized throughout the process of obtaining a policy and filing a claim.  

While the sophistication of the data utilized within the insurance industry has improved, so too has the scrutiny of how consumer data is requested and gathered, and how that data can be applied to better understand risk to write actuarially sound insurance rates.  

With new regulations adopted in Europe, new consumer data privacy laws being adopted at the state level in the US, and with no less than a half-dozen federal laws in the US regulating the use and disclosure of consumer data, it has become increasingly difficult for the insurance industry to provide new and innovative products and services, and set rates to an ever-expanding and demanding consumer base. 

The Rising Cost of Insurance Fraud  

Insurance fraud costs the economy, the insurance industry, and customers billions of dollars in economic damage every year. Criminals have been committing insurance fraud for as long as the concept of insurance has been around, and their schemes have evolved over the years. The use of consumer data and modern technology to steal personal identifiable information (PII), create synthetic identities, or hide from scrutiny has made fighting insurance fraud that much more difficult. With the increase in data sources and the growing sophistication of insurance fraud activity, insurers are now turning to artificial intelligence, machine learning, and algorithms to analyze consumer data in order to identify and prevent insurance fraud. However, in doing so, insurers must develop standards to prevent unfair discrimination and remove any bias from those technologies.  

Meeting Mandated Regulated Responsibilities 

In the same way that insurance companies should have a mandated responsibility to protect consumer data, they should be required by state law and regulations to identify, investigate, and report suspected insurance fraud. These two mandates are not mutually exclusive, and the insurance industry can protect consumer information and ethically utilized data sources to prevent, identify, and fight against insurance fraud and financial crime.  

Utilizing data and intelligence sources through permissive and public-record sources, and then ingesting that data into software models, not only assists in the prevention of insurance fraud, but it also allows companies to meet their mandatory regulatory responsibilities.  

The Importance of a Holistic View When Fighting Insurance Fraud  

Insurance fraud can occur throughout the process of applying, purchasing, using, selling, or underwriting an insurance policy. Drawing on the ability to accurately resolve and identify a true and holistic picture of bad actors puts the intelligence in the hands of the decision-makers. With the decision-making tools and customer intelligence in the hands of underwriters, claims adjusters, and investigators, insurance companies can better understand their customer needs, create new and innovative products, and process and fast track legitimate claims.  

With Consumer Data Comes Responsibility 

The utilization of artificial intelligence, machine-learning, and algorithms provides better visualization across legacy systems, as well as isolated and siloed industry data sources. These tools also improve risk assessment, customer intelligence, and processing capabilities across the entire company enterprise. However, with these types of technical capabilities comes a responsibility to protect and secure consumer data. Insurance companies must restrict unauthorized access, make sure protected consumer data is secure, develop meaningful policies and procedures, and continually review and analyze these standards in an open and ethical manner. 

As access and the use of consumer data increase, insurance companies must make data ethics a priority. Establishing an effective data security program with clear and transparent policies, procedures, use case standards, and an enforcement mechanism will provide a competitive advantage. 

Innovative technology should enhance and enable insurance industry professionals to make better, quicker, and more intelligent decisions. But this technology should never drive or replace the human decision-making process, but instead be used to improve the efficiency and effectiveness of activity.