French insurtech Shift Technology, a unicorn focusing on insurance fraud, was valued at over $1bn following its latest round of funding. The AI start-up offers insurance companies a set of software-based tools which can detect fraudulent behaviour across different scenarios, including claims fraud, underwriting fraud, subrogation detection, and financial crimes. The company also offers insurance partners a claims automation tool.
Currently, Shift Technology’s AI solutions are aimed at companies in the property and casualty sector. With the additional $220m it has raised in funding, the company hopes to expand in that sector while also entering the health insurance market.
Insurance fraud is a persistent issue that Shift wants to tackle. The FBI estimates the total cost of insurance fraud in the US to be more than $40bn a year. In the UK, the Association of British Insurers estimated that 107,000 fraudulent claims were made in 2019, worth around £1.2bn. The financial risk associated with fraud is high, and insurers must be increasingly vigilant as the tactics employed by fraudsters become more complex and harder to detect.
AI has the potential to accelerate fraud detection, with machine learning (ML) models able to filter through customer data and identify suspicious behaviour extremely quickly. Insurers can train their models on historical datasets to improve the accuracy of algorithms that detect fraudulent activity. Additionally, as insurers increasingly begin to automate their claims processes, integrating fraud detection capabilities will be a straightforward way of managing any risk associated with fraud.
Some insurance companies, including Axa and Liverpool Victoria, use data mining and ML-driven social network analysis to extract information from policyholders’ public social media activity. This allows them to glean further insight into the likelihood of a fraudulent claim.
As insurance services become increasingly digitised, the surface area for fraudulent attacks also grows. For example, minimising human employee involvement in the claims process via automation allows scammers to exploit and manipulate digital systems. Innovative solutions such as those developed by Shift Technology will be important in improving the methods insurers use to track fraud.
Not only can AI help tackle insurance fraud, but other financial services sectors can also use it to track and detect crimes such as money laundering and mule accounts. For AI vendors with fraud solutions aimed at one of these sectors, there is an opportunity to cross-sell and enter multiple markets.
Nonetheless, while AI will certainly improve and speed up fraud detection, ML algorithms need to be as fair and as accurate as possible. Minimising false positives in detecting fraud will ensure all customers are treated fairly and not discriminated against due to any bias in AI models.
Consumer trust in the financial services sector is already low. In 2020, communications firm Edelman identified the sector as the least trustworthy global industry in its annual Trust Barometer survey. Failing to pay out claims or withholding insurance coverage due to false fraud detection will further reduce trust levels amongst consumers.