In our review, we analysed the implementation of artificial intelligence in the insurance industry and how major insurance companies actually use artificial intelligence.

The insurance industry is one of the most data-driven sectors, utilizing data, statistics, and analytics. Big data, data mining, and processing technologies, combined with machine learning, have proven to be highly applicable in the insurance industry.

AI technologies began to be systematically applied in insurance around 2015-2016, and before the GenAI era, they already covered more than 70% of companies. In our analysis, we present how major insurance companies from the USA and the EU use AI in their processes, operations, and strategic decisions.

According to the Global Insurance Survey by Goldman Sachs in 2024, 29% of insurance companies are using AI globally (based on a poll of 359 CIOs & CFOs from insurance companies). Interestingly, according to the NIAC (National Association of Insurance Commissioners), 42% of insurance companies in America were already using AI in 2023.

So, the American market is leading in the adoption of AI and ML tools. Interestingly, according to the NIAC, insurance companies have already decided whether to use artificial intelligence and in what form. Thus, according to the NIAC, only 10% of companies were still considering the implementation of artificial intelligence, while 37% had already decided not to use artificial intelligence in their work.

We believe that due to the increased availability of GenAI-related products, companies have further expanded their use of AI, primarily in customer communications and internal document processing systems.

According to the NIAC, the majority of companies that have not yet implemented projects in the field of artificial intelligence stated that the implementation of the technology does not yet present compelling business reasons (56%). At the same time, 25% of companies indicated that they are still waiting for regulatory guidance.

The NAIC has developed the “Model Bulletin: Use Of Artificial Intelligence Systems By Insurers,” suggesting it as a set of temporary rules and a basis for future regulation. As of February 2025, this model has already been adopted in 21 states and is used in four states as a recommended document [current list of states]. This bulletin provides guidelines for insurers on the responsible deployment and utilization of AI technologies, emphasizing the need for comprehensive governance frameworks, risk management, and consumer transparency.

The model bulletin establishes uniform rules for enhancing consumer protection and reducing risks associated with AI deployment. This is a critical step that will protect innovation and the companies implementing it.

IBM survey of 1,000 global insurance companies and 4,700 insured customers notes that both companies and customers have significantly different views on generative AI expectations and concerns.

Insurance companies focus their AI efforts on core areas: pricing, marketing, risk management, and underwriting. At the same time, it is quite difficult to determine when the transition from machine learning technologies to the first models of artificial intelligence occurred.

We analyzed the largest European and American companies, studying how they use artificial intelligence and presenting the results of their utilization.

Artificial Intelligence in Insurance pricing

According to the NIAC, 54% of insurance companies already utilize artificial intelligence for pricing. Companies use credit and financial history, public records, demographics, driving behavior, and medical records to prepare personalized offers.

UnitedHealth Group (UNH) and Optum Bank back in 2017 developed a big data model for segmenting clients based on their behavior, enabling insurance companies to create more flexible insurance service packages. Effects: +26% in one-time contributions, +12% in average balances, and +23% increase in investment account openings.

Zurich Insurance Company and Quantee test real time personalised pricing models.

Anthem Insurance explore the use of AI in pricing, utilizing insights from medical history data and provider cost analysis to optimize the cost of services.

MunichRE reported implementing 45,000 models to estimate future spending and customer behavior, enabling it to deliver personalized offers and incentives.

Artificial Intelligence in Risk Management

Elevance Health and The Cigna Group (CI) implemented predictive models for chronic conditions to better manage pricing and alert customers to the need for early screening. (Sources: link 1, link 2)

CNP Assurances implemented the automatic analysis of health questionnaires. According to company data, this increased the automatic acceptance rate for the benefit of policyholders by 5% (reaching more than 80%).

AXA XL enhanced property risk engineering capabilities by automating the process of data extraction from site survey reports. The AI accelerates the analysis of more than 10,000 such reports every year.

Allstate, AXA XL, Farmers, MetDirect, Berkshire Hathaway introduced predictive analyses of wildfires using geospatial imagery and risk factors, including topography and the proximity of electrical cables to buildings. (Sources: link 1, link 2)

AXA XL introduced the AI Digital Risk Engineer to monitor the health of assets and buildings by capturing and intelligently processing data from connected systems such as heating, ventilation, and air conditioning.

Artificial Intelligence in Underwriting

According to the NIAC, 66% of insurance companies that already utilize AI use it for approval and denial cases.

Allianz introduced the ‘Incognito' system for fraud detection in motor claims, home insurance claims, and applications. The system analyzes distortions in real-life images, videos, and documents. According to company estimates, it led to a 29% increase in fraud detection. Zurich Insurance Company also utilizes AI for fraud detection in claims processing.

Aviva uses AI for claims operations to improve outcomes at every step of the process. The company announced that AI helps assess liability for complex cases 23 days faster and has improved routing accuracy by 30%. This helped Aviva reduce customer complaints by 65% and save £100 million.

Aviva, in collaboration with Tractable, announced the use of photos to assess the damage to housing and automobiles, allowing for accurate remote assessment of repair estimates and increased consistency in diagnosis.

Artificial Intelligence in Communications & Marketing

CNP Assurances enhanced customer interactions by introducing document classification to speed up customer procedures by identifying, checking, and classifying documents.

Crédit Agricole Assurances AI models provide personalized advice by analysing customer data to identify coverage needs and suggest appropriate insurance products.

Humana Inc., together with IBM, introduced an AI-enabled virtual agent designed to provide members with accurate and personalized information regarding benefits, coverage, claims, referrals, and healthcare costs for both private and corporate customers.

MetLife & Nayya utilize AI for customer decision support, proactive notifications, and claims-based opportunities.

Allstate implemented AI in its customer claims response system (handling about 50,000 messages per day). The company announced that the use of AI made responses more accurate, empathetic, and unique, while effectively handling insurance jargon.

‍

‍

Sources:

[1] Global Insurance Survey, 2024. (link)

[2] Life Insurance Artificial Intelligence/Machine Learning Survey Results, NAIC Staff Report, 2023. (link)

‍

‍

‍

Tags: #pricing #underwriting#insurance #artificialintelligence #AI #communications #marketing #UnitedHealthGroup#Quantee #AnthemInsurance #MunichRE #ElevanceHealth #Cigna #AXAXL #Farmers #MetDirect#BerkshireHathaway #Allianz #ZurichInsuranceCompany #Aviva #Tractable #CNP #CréditAgricole#Humana #MetLife #Allstate #GlobalInsuranceSurvey #NAIC