Insurance – Personality Assessment for Fraud Detection

Problem

 Fraud detection in Insurance

The Coalition Against Insurance Fraud estimates that illegitimate insurance claims cost about USD 80 billion every year. At that, 10 per cent of people consider insurance fraud to be a victimless crime. However, fraudulent claims raise the price of insurance for everyone and thus generate significant expenses. It is obvious that claim legitimacy check and Personality Assessment for Fraud Detection are in the best interest of both insurance companies and diligent insured persons.

An insurance company organizes investigation for evaluation of a claim legitimacy. After, it uses the obtained investigation results to combat the prevalence of false or inflated claims and avoid paying potentially significant costs to fraudsters. The main types of investigated claims are the following:

Evaluation of a claim legitimacy in insurance

1. Workers’ Compensation Claims

The investigation here seeks to determine two things:

  • Is the employee as injured as they claim to be?
  • Was the injury acquired while the person was working?

2. Personal Injury Claims

Personal injury claims can be filed against either a business or against another person. For example, such a claim becomes fraudulent when the victim actually fell on their own icy steps but staged the incident to look like it occurred in front of a company’s storefront.

3. Property Damage and Theft Claims

In order to investigate property damage (e.g., fire damage, water damage or car accidents) and theft claims (e.g., theft, burglary, hijacking or robbery), the relevant industry experts may be involved. The information gained through this investigation process helps the examiner either to confirm or deny that the claim is legitimate.

4. Healthcare/Medical Fraud Claims

Both a practitioner and a patient, sometimes together, can participate in fake or inflated healthcare claims to line their own pockets. According to the Legal Information Institute, 10 cents of every dollar spent on healthcare goes toward paying for fraudulent healthcare claims.

An insurer is entitled to investigate the claim to make sure it is genuine and falls within the scope of the correspondent policy. Backing for decision-making is represented by collected evidences, records, and interviews. The final conclusion states if the claim under investigation is legitimate or illegitimate. In many cases it would be extremely useful to have a reliable personality portrait of an insured person to make this final conclusion even more reliable. Unfortunately, a dedicated psychological testing procedure is usually not provided for. However, Sentino has a smart solution that is able to bring prominent personality insights.

Solution

Sentino Interview Semantic Analysis API as a psychological basis for claim verification and fraud detection

Sentino proposes Interview Semantic Analysis API, an AI-powered tool capable to generate comprehensive personality portraits based on processing of audio or video records of interviews. Investigators can save time and effort, since no additional actions are required on their side. Instead, they can file an interview record to Sentino and in a short while receive a multifaceted psychological portrait of the insured person subject to the investigation.

The whole personality assessment process includes just 6 easy steps:

1. Determination of personality traits of interest

Determination of personality traits and facets of the insured person, which are of particular interest for investigators and require assessment. Currently, there is a list of up to 250 traits available. They are determined based on 20+ validated and widely used psychological inventories (Big 5, NEO-PI, RIASEC, DISC, ORVIS, etc. – the list can be further supplemented by the request).

2. Interview record submission

Uploading of an audio or video file via Sentino Interview Dashboard.

3. Interview record pre-processing

Conversion of audio or video record into text with the use of AI-based transcription software.

4. AI-powered interview scoring

Semantic analysis of the interview text performed by Sentino API and generation of the comprehensive personality portrait that includes the traits selected within Step 1.

Interview Semantic Analysis API is a data-driven AI tool developed based on the state-of-the-art NLP/ML technologies. We continuously perform researches and upgrade our technical solution striving for high accuracy. Our API transforms the unstructured natural text of the interview into psychologically meaningful representation and links it to scientific personality inventories.

5. Report downloading

Downloading of the complete report from Sentino Interview Dashboard.

6. Decision making

Now the final decision can be made in relation to the claim legitimacy. It is properly substantiated by previously collected facts and evidences from different sources, as well as by the insightful personality portrait.

Please note that ethical approach to AI employment requires adhering to fundamental values like individual rights, privacy, non-discrimination, and non-manipulation. That is why AI should not be used as a single decision-making source. Best practices stipulate its application in conjunction with other sources of information and human judgement. Ethical use of AI opens up numerous and significant advantages.

Benefits

  • Time and cost saving
  • Standardized scientific approach to claim verification
  • Validated and globally recognized assessment tools
  • Careful consideration of specific personality traits
  • Fair and unbiased decisions
Sentino API - Insurance benefits
Scroll to top