How AI and advanced analytics can benefit policyholders

Artificial intelligence (AI) carries enormous transformational potential for industries and society. Its benefits have been widely recognised and it has become an area of strategic importance for the EU and a key driver of economic development. By harnessing the power of AI, insurers can improve efficiency gains and better serve their customers, while also promoting transparency and fairness.

The potential benefits of AI in insurance are numerous, particularly for consumers. AI can be used to develop insurance products and services that are better tailored to customers’ needs and preferences, while also helping insurers to better understand and mitigate potential risks, and to reduce the likelihood of claims.

All together, this can lead to better outcomes for policyholders, reducing the risk of underinsurance or overinsurance, and contributing to significantly reducing the insurance protection gap.

Insurance Europe has held a series of interviews with experts from European insurers to gather their views on current and anticipated uses of AI in insurance. The results of these interviews have informed the content of this document.

This document explores how insurers can leverage AI and advanced analytics to enhance efficiency gains, improve customer service and prevent fraudulent activity, looking, for example, at motor and natural catastrophes. By leveraging the potential of AI while ensuring ethical and regulatory compliance, insurers can drive positive change and better serve the needs of consumers.

Use case: Helping in the management of material motor claims

For eligible claims and on the decision of the claims manager, in agreement with the insured, AI machine learning analyses photos of the insured’s vehicle and can propose either:

a recommendation to repair or replace certain vehicle body parts before final decision by an experienced claims assessor; or,
compensation to the insured, which they may accept or refuse.

The insured always retains the right to request appraisal by a human expert and automobile experts’ assistance on the damage caused.

Benefits

No need for a physical appointment to assess the damages on-site, streamlining the claim assessment process and reducing its time and cost.
Reduced time to make the reimbursement.

Use case: Helping in the management of material motor claims

For eligible claims and on the decision of the claims manager, in agreement with the insured, AI machine learning analyses photos of the insured’s vehicle and can propose either:

a recommendation to repair or replace certain vehicle body parts before final decision by an experienced claims assessor; or,
compensation to the insured, which they may accept or refuse.

The insured always retains the right to request appraisal by a human expert and automobile experts’ assistance on the damage caused.

Benefits

No need for a physical appointment to assess the damages on-site, streamlining the claim assessment process and reducing its time and cost.
Reduced time to make the reimbursement.

Importantly, AI systems also enable a shift from current damage indemnification models towards risk mitigation and prevention, which provide more value to customers than simple reimbursement.

The development of AI tools can help insurers to better monitor and predict risk, enabling them to advise policyholders on how to reduce risk, which can in turn help reduce the frequency and severity of losses over time. For example, data from sensors, telematics devices in vehicles or smart devices in homes and factories can anticipate and potentially prevent harm to property, machines, vehicles and people.

Importantly, AI systems also enable a shift from current damage indemnification models towards risk mitigation and prevention, which provide more value to customers than simple reimbursement.

Deep dive: a wide variety of potential benefits

Examples of monetary and qualitative benefits to consumers of AI applications across the insurance value chain

AI can add value for customers in a variety of both monetary and qualitative ways. The latter includes essential factors such as increased insurability, greater ease in initiating a contract or settling a claim and a general feeling of being treated fairly.

Examples

AI applications in insurance cover a wide range of customers’ demands and needs

The range of applications in which AI has considerable potential to positively impact policyholders spans healthcare, motor, fire and multiple lines — a non-exhaustive list is in the table below.

Examples

AI applications in insurance cover a wide range of customers’ demands and needs

The range of applications in which AI has considerable potential to positively impact policyholders spans healthcare, motor, fire and multiple lines — a non-exhaustive list is in the table below.

AI applications in insurance cover a wide range of customers’ demands and needs

The range of applications in which AI has considerable potential to positively impact policyholders spans healthcare, motor, fire and multiple lines — a non-exhaustive list is in the table below.

Examples

Looking ahead, generative-AI-powered tools, such as advanced voice recognition and text generation, could help streamline the customer experience when processing paperwork and digital forms. Additionally, they might help with answering policyholders’ questions about contracts and terms and conditions more efficiently.

Examples

Why AI applications can improve insurability, fairness and consumer inclusion

AI applications can address demand-side barriers to purchasing insurance coverage, drive enhanced coverage of risks (eg, for natural catastrophe events), increase the insurability of events that might result in damage, make it possible to cover more consumers and help to reduce protection gaps¹.

Addressing protection gaps is an issue not only for individuals but also for society. For example, in a natural catastrophe, only around half of European consumers could be fully or partially covered.

For natural catastrophe events, insurers are developing technological capabilities to assess risks accurately, respond more quickly to disasters and educate policyholders on effective prevention measures and risk management practices.

These advances lead to innovative ways to underwrite insurance, allowing for enhanced risk measurement and increased acceptance of consumers’ applications for cover.

AI applications can use historical data on natural disasters (eg, forest fires, floods, tornadoes, etc.) and disaster images transmitted by the insured to predict the occurrence of a natural event (eg, spread of a forest fire, likelihood of flooding) and analyse the origin of the damage (eg, a crack linked to a structural defect in a house or to drought in the ground beneath it).

Insurers are integrating multiple approaches and expertise from various scientific disciplines to build their understanding of natural catastrophe events, enhance prediction capabilities and expedite claims processes:

Aerial imagery and AI: Specialised AI processors and aerial inspection programmes to expedite the identification of heavily impacted areas after storms².
AI imagery: AI solutions help homeowners assess damage and recover from typhoons and expedite the claims process³.
Data analytics and AI: Services to enhance resilience to climate change by identifying specific exposure to regional events that might result in property damage and estimate losses⁴.
AI and risk modelling: A roof-shape classification AI model helps to define customers’ insurance coverage needs specific to their roof type⁵.

Benefits

Helps to prevent harm to people and property, reducing potential damages and related costs.
Helps to identify and call in experts.
Eases and speeds up claims settlement processes.

Benefits

Helps to prevent harm to people and property, reducing potential damages and related costs.
Helps to identify and call in experts.
Eases and speeds up claims settlement processes.

Conclusions

Regulatory and disclosure requirements have accumulated over the years, making application processes more cumbersome for both customers and insurers due to laborious paperwork and lengthier forms.

AI applications in the insurance sector are already giving rise to efficiency gains, helping to improve customer service and offering more insight into customers’ needs. The potential benefits of AI applications in the insurance sector are numerous, including:

processing claims faster or even instantly;
Providing more tailored services and more relevant coverage
Improving risks prevention
Enhancing risk insurability

Conclusions

Regulatory and disclosure requirements have accumulated over the years, making application processes more cumbersome for both customers and insurers due to laborious paperwork and lengthier forms.

processing claims faster or even instantly;
Providing more tailored services and more relevant coverage
Improving risks prevention
Enhancing risk insurability

Advanced analytics provide new opportunities to create added-value for customers by identifying, assessing and managing risks to reduce the protection gap, offer better coverage to vulnerable people and reduce costs for policyholders and societies. The application of AI can also help to enhance insurability and allow insurers to provide cover against specific risks that were previously uninsurable.

As the insurance industry continues to evolve, it is crucial to leverage the latest technology to create beneficial outcomes for consumers.

1. A protection gap is the difference between the amount of insurance that is economically beneficial and the amount of coverage actually purchased

2. “Speedy recovery: How AI research is helping insurers boost claims response”, M. Chaban, Transform, September 2022

3. “New AI solution to help homeowners recover faster from natural disasters”, Tractable, January 2022

4. “Zurich launches dedicated Climate Change Resilience Services to help businesses tackle climate change risks”, Zurich Insurance Group, September 2020

Value-chain component	Monetary value/tangible value	Qualitative value/intangible value
Product design & development	Better alignment between product features and target audience can reduce the overall cost for the insured, both directly (eg, excessive coverage costs) and indirectly (eg, reduced cost of insurer’s capital and maintenance costs over multi-year periods).	More flexible and relevant products. Greater customer satisfaction and longer-term relationships.
Pricing & underwriting	Better risk-based pricing without subsidising non-distinguishable risk. Cover for previously uninsurable risks (and reduced community costs).	Better insight into the actual price of risk can make previously uninsurable risks insurable. Process efficiency means the cover can be bought online and concluded in real time.
Sales & distribution	Improved, proactive information on more relevant products, reducing purchases of unnecessary cover. Lower cost of distribution through automation.	Less time is needed to get to the right product. More competitive prices.
Customer service	Process efficiency can reduce costs.	More consistent service quality, shorter waiting periods and more flexible ways to interact with the insurer.
Loss prevention	Reduced claims levels allow the insurer to lower the cost of loss in the pricing.	Coupling AI with Internet of Things (IoT) reduces damage and risk events through active prevention.
Claims management	Reduced levels of fraud allow an insurer to lower the cost of fraud included in its pricing, thus reducing the price for non-fraudulent customers. Lower cost of processing claims thanks to simplified processing.	Shorter response time from the first notice of loss to claim settlement.

AI Applications	Description	Benefit to policyholders
Health: conversational AI	Provides support throughout end-to-end physiotherapy rehabilitation. A virtual agent offers consultations and recommendations to enhance and accelerate recovery from musculoskeletal injuries.	Improved customer experience and reduced time and cost of involving physiotherapy experts. Reduced costs from fewer sessions necessary for full recovery.
Motor: real-time accident support	Insurers can deliver superior service levels to drivers by receiving automatic access to accident data and providing a rapid, semi-automated response.	AI chatbot can prompt the driver on the best actions for recovery, automatically notifying the medical team if needed or calling a tow-truck service.
Motor: education	Use sensors to improve customers’ driving and collect data. Timely feedback helps to improve personal driving safety.	Receipt of messages to improve driving and reduce risk.
Fire	Predictive models provide a score that estimates how at-risk a given property is to fire.	Fairer pricing and a need for less information collection.
Multiple lines: digital claims management	Online processing of first notices of loss.	Improved customer satisfaction. Faster claims payment.
Multiple lines: fraud prevention in claims processing	Using AI systems for fraud detection or to control claims costs enables insurers to keep claims costs under control.	The majority of honest clients are supported through a reduction in costs and premiums.
Multiple lines: client proposition	Suggests relevant and valuable products for individual customers (in compliance with the proper demands and needs assessment under the Insurance Distribution Regulation.	Tailored services, more relevant propositions.
Multiple lines: voice server	Interaction with a computer-operated telephone system.	Requests are handled more efficiently. Reduction in waiting time and improved experience.
Multiple lines: predictive risk modelling	Enhance risk prevention (educate, inform, improve).	Loss prevention can reduce insurance premiums. Improved customer experience.