Could voice-based risk assessments be the answer?

Written by: Manjit Rana, Executive VP, Insurance at Clearspeed

In Britain, around 53 million people own a smartphone. That’s 95% of the population, rising to 98-99% when accounting for individuals aged 16-54. With such widespread adoption, and flagship smartphones costing more than £1,200, many individuals are turning to insurance to protect their mobile devices.

Resultantly, gadget insurance policies are on the up, rising from 7.87 million in 2023 to 8.46 million in 2024, with Gen Z being a key growth demographic.

While gadget insurance is often the first policy that young customers will buy, these same customers expect their insurance experiences to be seamless, transparent and digital, much like other products and online services they use every day.

To build trust, insurers must meet these demands with modern customer and user experiences. Yet those experiences cannot come at the expense of proper claims scrutiny.

In the UK, gadget insurance gross written premium income increased from £496 million in 2023 to approximately £604 million in 2024. A claims frequency of 7.8% was also recorded in 2024, meaning that approximately 660,000 claims were made during the 12-month period.

However, it’s estimated that as many as 5-15% of UK gadget insurance claims could be fraudulent, meaning that as many as 99,000 falsified, deceptive or misleading claims could have occurred in 2024 alone. With the UK market seeing typical payouts of £435 per claim, this may have translated to a financial impact of £43 million.

Despite the huge costs, rooting out and eradicating fraudulent claims isn’t an easy task at present. ‘Lost’ claims, for example, don’t generally require police reports, with many claimants filing stolen or damaged devices as lost to avoid additional checks.

Insurers will often require proof of purchase receipts with any device claim, but the reliability of those evaluations is decreasing. AI can now be used to fabricate or amend invoices and receipts, for example. And the presence of receipts does not ensure claimants won’t misrepresent how or when a device was lost or damaged.

It’s a difficult dilemma for insurers. Customers – particularly Gen Z policyholders – are increasingly demanding easy claims processes and rapid resolutions, particularly given the importance of mobile phones and other gadgets to daily life. Yet insurers cannot afford to relax fraud detection efforts, many of which are slow, can be bypassed by opportunistic fraud, and are resource intensive.

Could voice-based risk assessments be the answer?

There are some additional technologies already in place that can help insurers in combatting fraud.

The Recipero database enables insurers to check International Mobile Equipment Identity (IMEI) codes against databases to see if devices have been sold or recycled. They can also cross-check claims against other insurance databases to identify potential duplicates – a common indicator of fraud.

Network data requests can help insurers to see when a phone was last used to validate claimed loss dates, while exclusion periods can prevent customers from purchasing policies and making claims immediately afterwards.

However, these methods too have their flaws. Timing gaps, for example, can be exploited by taking out policies and claiming losses shortly after exclusion periods conclude.

In gadget insurance, and the wider industry, trust is the currency of successful policy-carrier relationships. The insurer needs to be confident that the consumer is honest and accurate with the information they’re providing, and the claims they’re submitting. The consumer, meanwhile, needs to be confident that their claims will be handled with speed and fairness.

Right now, that insurer’s trust is being undermined by the high volume of fraudulent claims, and consumer trust is being impacted by the laboriousness of legacy checks. However, the trust gap can be bridged with new methods and innovations, such as AI-assisted routing, document verification tools, and risk assessment solutions.

Take voice-based risk assessment as an example. The vocal characteristics associated with risk are universal across language, age and other demographics. By asking simple yes or no questions, insurers can quickly identify low-risk customers, allowing straightforward claims to move faster while reserving deeper investigations for those cases that warrant additional scrutiny.

New technologies depend on responsible usage

In adopting modern innovations such as these, legacy checks can be overhauled with more suitable solutions for detecting fraudulent behaviours, all while meeting Gen Z demands for rapid claims resolutions.

Part of the effectiveness of this approach is that it not only effectively helps in indicating fraud, but it also prompts individuals to think twice before misrepresenting claims.

Of course, these technologies must be used and implemented in a responsible and compliant manner. For example, they should not be used as a sole determinant of fraud, but as a risk triage tool that facilitates the clearing of most cases and focused followed ups. In an industry that is stringently regulated, and heavily reliant on decision accuracy and defensibility, this proper use if particularly important.

Further, any solutions must also align with the FCA’s Consumer Duty, which requires regulated companies to put in place controls to protect customer data, or prevent fraud from arising from the misuse of personally identifiable information (PII).

Ultimately, the challenge for insurers is balancing fraud identification with Gen Z expectations around speed, convenience and fairness in gadget insurance. By replacing laborious, time-consuming, costly,

ineffective and frustrating legacy checks with new innovative means of building and sustaining customer trust while mitigating fraud, both insurers and their policyholders will benefit.

Image provided by Clearspeed

Alicia Ward