Artificial intelligence applications in health insurances: a scoping review

Abstract

Introduction

The rapid evolution of technology has reshaped the insurance industry, with artificial intelligence (AI) taking center stage as a key driver of innovation. This paper examines the transformative impact of AI in health insurance, focusing on its applications and potential to revolutionize the sector.

Method

This scoping review examines literature published between 2000 and 2024, focusing on the application of AI in health insurance. We used relevant keywords related to artificial intelligence and health insurance to search the PubMed, Scopus, and Web of Science databases.

Findings

AI presents numerous opportunities in health insurance, including contributions to shaping international and national agendas, such as aligning goals, establishing indicators, and achieving objectives, financial management, fraud detection, monitoring capabilities, diagnostics and medical innovations, private insurance applications, risk management, technical analysis, and value creation. However, there are ethical challenges that must be addressed if AI is to be effectively implemented.

Conclusion

Policies for AI applications in health insurance should prioritize the protection of personal health and medical data, address ethical concerns, and ensure robust data privacy and security. Additionally, these policies should promote the use of AI to enhance customer experiences, optimize risk selection, and generate revenue for both insurers and policyholders.

Introduction

Insurance plays a pivotal role in protecting individuals and organizations by mitigating the financial risks associated with death, illness, property loss, and more [1, 2]. However, the growing imbalance in medical insurance revenues and expenditures, coupled with escalating healthcare costs, has created an urgent need for effective risk control measures. To address these challenges, innovative solutions such as AI-driven systems are emerging as vital tools for ensuring the financial sustainability of medical insurance funds [3].

Advances in artificial intelligence (AI) and big data analytics are transforming health insurance by enabling accurate cost predictions, fraud detection, and optimized resource allocation. Algorithms developed for insurance risk scoring have significantly improved reliability, allowing insurers to forecast health outcomes and refine premium pricing [4]. AI-powered virtual health assistants further enhance healthcare accessibility by offering personalized, real-time support, automating routine tasks, and improving adherence to treatment plans. These tools provide multilingual, 24/7 assistance, particularly benefiting underserved populations while enhancing overall patient care [5].

By 2025, the proliferation of one trillion connected consumer devices, along with innovations such as autonomous drones and 3-D printing, is expected to revolutionize risk management, enabling real-time assessments and operational efficiency. By 2030, AI will fully transition the insurance industry from a “detect and repair” model to a “predict and prevent” paradigm, leveraging technologies like smart contracts, adaptive AI models, and instant underwriting to make insurance more dynamic and customer-centered [6].

While these advancements underscore AI’s transformative potential, previous reviews on AI in health insurance have often been narrow in scope, focusing primarily on specific applications such as fraud detection or cost reduction. Critical areas, including the integration of clinical data with insurance systems, the role of cost-sharing mechanisms in ensuring equitable healthcare, and AI’s broader systemic impacts, remain underexplored. Additionally, few studies have consolidated these elements into a cohesive framework that addresses the growing complexities of healthcare systems and insurance processes. This gap highlights the necessity for a comprehensive examination of AI’s contributions to the insurance sector [7, 8]. The incorporation of AI into healthcare has already demonstrated significant benefits by streamlining medical tasks, improving diagnostic accuracy, and facilitating more effective treatments. By rapidly processing vast data volumes, AI reduces diagnosis and treatment times, offering notable advantages to both healthcare providers and patients [1, 2]. AI is enhancing various aspects of the insurance process, from analyzing bills and expenditures to determining appropriate pricing for medical services, contributing to sustainable healthcare provision [8–10]. The integration of clinical and insurance databases strengthens AI’s capacity to improve decision-making and operational efficiency, attracting interest from institutions and government agencies alike [11]. Moreover, AI facilitates equitable resource allocation by enabling cost-sharing systems that distribute healthcare costs more efficiently while ensuring patients receive the care they need without facing undue financial hardship [9, 10].

With its numerous interconnected subsystems and massive data sets, the healthcare industry stands to gain significantly from AI applications in health insurance [12]. AI can streamline various processes, enhance data analysis, and improve decision-making, leading to more efficient and effective healthcare delivery. Moreover, the function of cost-sharing systems further emphasizes the importance of AI applications in health insurance [13, 14]. Despite the extensive applications and benefits of AI in health insurance, comprehensive studies on the subject are scarce in the literature. While AI has demonstrated its potential across various domains, research specifically focused on its applications within the health insurance sector remains limited. This scoping review aims to address this gap by systematically exploring the existing literature, highlighting the transformative potential of AI in health insurance, and identifying areas for future research.

Methodology

We followed the scoping review framework proposed by Arksey and O’Malley [15], involved five key stages:

Formulating the research question

The primary research question guiding this review was: “What are the applications of AI in health insurance?” This question was designed to explore the dimensions, and areas of AI applications within health insurance systems.

Identifying relevant studies

A comprehensive search strategy was developed and executed across three major databases—PubMed, Web of Science, and Scopus—to ensure broad coverage of relevant studies. The search strategy incorporated a combination of controlled vocabulary and free-text keywords related to “artificial intelligence,” “health insurance,” “machine learning,” “big data,” and “intelligent systems.” Boolean operators such as AND/OR were used to combine keywords effectively. Searches were restricted to articles published between 2000 and 2024, allowing for the inclusion of both foundational and contemporary research. Table 1 provides the complete search queries, illustrating the exact keyword combinations tailored for each database. Duplicate records were removed systematically during the search process to maintain accuracy.

Table 1.

Search queries used for target databases

Databases	Query	Results
PubMed	(health[Title]) AND (Insurance[Title]) AND ((“artificial intelligence“[Title] OR “machine learning“[Title] OR “big data“[Title] OR “internet of things“[Title] OR “expert system“[Title] OR “intelligent system“[Title] OR “Knowledge system“[Title] OR “data mining“[Title] OR “deep learning“[Title]))	52
Web of sciences	TI = (“artificial intelligence”) OR TI = (“machine learning”) OR TI = (“big data”) OR TI = (“internet of things”) OR TI = (“expert system”) OR TI = (“intelligent system”) OR TI = (“knowledge system”) OR TI = (“data mining”) OR TI = (“deep learning”) (TI = (insurance)) AND AB = (health)	65
Scopus	((TITLE ( "artificial intelligence" ) OR TITLE ( "machine learning" ) OR TITLE ( "big data" ) OR TITLE ( "internet of things" ) OR TITLE ( "expert system" ) OR TITLE ( "intelligent system" ) OR TITLE ( "knowledge system" ) OR TITLE ( "data mining" ) OR TITLE ( "deep learning" ))) AND (TITLE ( "insurance" )) AND (TITLE ( "health" ))	108

Selecting the studies inclusion criteria

This study utilized carefully defined inclusion criteria to ensure the selection of relevant and high-quality articles. Only studies specifically addressing the intersection of AI technologies and health insurance were considered. Furthermore, the review was restricted to publications written in English and released between 2000 and 2024, guaranteeing linguistic accessibility and contemporary relevance.

Exclusion criteria

To ensure the selection of studies was both rigorous and aligned with the objectives of this research, specific exclusion criteria were applied. Articles that did not focus on health insurance or AI applications were excluded from the review. Additionally, studies lacking full-text availability or failing to meet acceptable methodological standards were omitted. This process helped maintain the integrity and relevance of the selected literature.

Title and Abstract Screening: Conducted independently by two reviewers to filter out irrelevant studies.

Full-text evaluation

In-depth review by the same two reviewers to ensure alignment with the inclusion criteria. Discrepancies were resolved through discussion, with input from a third reviewer when necessary. A detailed screening process is visualized in Fig. 1 (flowchart).

Fig. 1.

Flow chart of the search strategy

Charting the data

wo authors (MR and AB) independently coded the data to ensure accuracy, and discrepancies were resolved through team discussions. The analysis culminated in a conceptual framework (Fig. 2) that categorizes and explains AI applications in health insurance.

Fig. 2.

AI applications and capabilities for health insurances (contemporary studies and future agenda)

Collating, summarizing, and disseminating results

The final stage synthesized the data into a narrative using systematic, transparent narrative synthesis techniques. This synthesis was designed to contextualize and consolidate key findings. The research team, led by the corresponding author (AT), reviewed and approved the narrative. Results are presented in two components: 1- A structured analysis of the findings, providing an overview of the themes and patterns identified. 2- A narrative discussion that contextualizes the implications of the findings within the broader context of AI applications in health insurance.

Results and discussion

Our initial search identified 225 references, 83 of which were duplicated. We screened a total of 142 articles with title reviews, which resulted the exclusion of 38 articles. 104 studies met the inclusion criteria for full-text evaluation, and 44 studies were excluded due to the lack of relevancy or poor methodological quality.

The results are presented in two parts: a summary of the extracted information (Appendix 1); a proposed framework to explain its dimensions (Fig. 2).

A framework for future studies: existing knowledge and future directions

The detailed explanations of each dimension—International and National Agendas, Financial Aspects of Health Insurance, Monitoring Capabilities, Diagnosis and Medical Revolutions, Private Insurance Applications, Risk Management, Technical Analysis, Value Creation—are provided to clarify their specific roles and contributions. Additionally, Ethical Challenges and Responsibilities, as highlighted in previous studies, are also included in the paper to ensure a comprehensive discussion.

International and national agendas (goals, indicators, and objectives)

Machine learning (ML) aids policy decisions by providing causal estimates and evaluating counterfactual outcomes, helping policymakers design effective health insurance policies [16]. It enables the development of customized insurance plans based on customer data, allowing insurance agents to offer tailored payment options, coverage amounts, and policy durations [17]. AI tools can analyze various aspects of the insurance sector, including service pricing, sales, delivery, cost management, claims management, fraud detection, and the creation of services for insured individuals [14]. An AI model was developed to predict health insurance premiums by considering factors such as age, gender, BMI, number of children, smoking habits, and geography, demonstrating improved performance in predictive accuracy [8].

Governmental cooperation is necessary to establish a healthcare big data standard and specification system, ensuring data integrity, authenticity, and reliability. AI also plays a critical role in health management services and pharmacy benefit management (PBM). Health management services use AI to track and evaluate health, prevent diseases, reduce expenses, and predict complications. PBM relies on healthcare big data for real-time prescription audits and settlement reconciliation [18]. AI significantly impacts health insurance by accelerating drug development, allowing scientists to focus on creating relevant treatments. This efficient use of time and resources can be life-saving in emergencies. AI can also analyze data from fitness trackers and wearable devices to assess health and determine risk profiles, enabling insurers to offer personalized and accurate policies and streamline the underwriting process. Additionally, AI-powered chatbots provide instant and accurate information to customers, improving service and reducing the need for human interaction. Overall, AI in health insurance leads to cost savings, improved efficiency, and better customer experiences [8].

Financial aspects of health insurance

AI and machine learning are transforming health insurance by improving pricing accuracy and policy design. ML models outperform traditional actuarial models in predictive accuracy, leading to cost reductions and better pricing strategies. These models also identify new concession opportunities, helping underwriters’ price and retain policies more effectively [19]. For instance, one study focused on minimizing weight deductions for hospitals and improving the appeals process for claim denials. By analyzing health insurance claims with big data analytics, the study developed a predictive model that reduced deductions and achieved significant cost savings [20]. AI tools also play a crucial role in promoting healthy habits, e.g., exercise and a healthy diet, which can reduce healthcare costs [8]. Furthermore, addressing higher mortality rates and costs associated with managing HIV, AI can help reduce comorbidities, address lifestyle factors, and screen for cancers and cardiovascular diseases, ultimately lowering healthcare expenses for individuals living with HIV [9].

AI can also tackle inequity and reduce financial burdens [21]. By optimizing co-insurance rates based on drug costs, expenses, and insured characteristics, leading to potential cost savings [13]. Another study analyzed the prevalence and economic burden of congenital cytomegalovirus (CMV) in Korea using big data from the national health insurance system, examining the average mean healthcare expenditure per capita [10]. Moreover, AI helps understand optimal health insurance structures by applying Gaussian process priors and analyzing the trade-off between insurance/redistribution and provider expenses, highlighting the importance of healthcare consumers’ behavior in response to changes in the coinsurance rate [16]. AI tools are also vital in payment methods and cost management. Diagnosis-Related Groups (DRGs) and Disease Intervention Programs (DIP) standardize hospital payments and evaluate the correlation between diseases and treatments. These methods rely on healthcare big data to monitor expenses and promote efficient, low-cost healthcare, addressing over-servicing and encouraging competition among hospitals [18]. Additionally, a study proposed using a hybrid model (HM) approach to assist in DRG policymaking, combining automated methods and statistical knowledge. This approach helps extract useful information from claims data for further assessment and addresses concerns about excluding expensive medical materials from DRG calculations [22].

Fraud management

Researchers are developing AI systems to assess health insurance claims and detect fraud, which can help insurers settle legitimate claims faster. Manual processes and fraudulent claims cause delays in claim processing. AI can reduce processing times by identifying fraudulent claims and learning from past data, greatly improving the efficiency of health insurance claim processing [8, 23]. Data mining techniques are used to analyze healthcare insurance data and identify fraudulent behaviors, which negatively impact businesses in various ways [24]. One study recommended using fuzzy expert systems with AI to detect fraud in health insurance, improving accuracy in identifying fraudulent cases [25]. Another study developed an AI-based model to detect fraud in health insurance claims in Saudi Arabia, accurately identifying the main factors contributing to fraud [26]. Additionally, an AI-driven tool using big data was developed to detect and forecast fraud in health insurance, successfully identifying fraudulent activities such as self-referral, collaboration with providers, and double billing [24].

AI is also used to address fraud, waste, and abuse (FWA) in health insurance. Companies have deployed AI-driven solutions to facilitate FWA identification. For example, Kirontech’s CaseHound software uses public health data to train AI models, while Azati’s platform identifies unusual claims based on previous data. These AI-powered solutions aim to reduce FWA in the health insurance industry [7]. Another study highlighted the application of AI in health insurance claims to improve the verification process and minimize miscarriages of justice. The study proposed using data-mining technology to identify cases requiring manual inspection, saving time and resources. Integrating data-mining technology into different stages of the business process and implementing an early warning system are important for the future advancement of health insurance systems [27]. Traditional and advanced statistical data analysis tools and algorithms are used to predict fraudulent claims in health insurance, with the primary aim of developing a methodology that achieves business goals in healthcare fraud management [28]. Moreover, another study introduced a claim prediction system using machine learning techniques to assist auditors in correcting claims. This system incorporates feature selection, concept drift, and active learning to gather feedback from auditors and improve over time, and has been applied to two large US health insurance companies [29]. AI in health insurance can improve transparency, data security, and customer privacy, thus eliminating discrimination and ensuring legal justice [18, 30, 31]. Developing a mortality model and life-scoring tool using large datasets can reduce claims by 9% in the healthiest applicants [32].

Furthermore, responsible AI (RAI) can help hospital administrators identify potentially denied claims, thereby improving profitability, accelerating revenue cycles, and supporting patient well-being by reducing operational costs and increasing efficiency [33].

Monitoring capabilities

AI tools can provide a better understanding of the diverse impacts of health insurance programs, especially for populations with limited access to healthcare. One study proposed using causal machine learning to estimate individual treatment effects and develop optimal policy rules, targeting areas where the greatest benefits can be attained [34]. Another study highlighted AI’s role in improving the administrative review process of medical expenses by enhancing data classification accuracy, thus improving medical cost audits and reducing resource misallocation [35]. Furthermore, AI tools have been used to evaluate the impact of different health insurance schemes on maternal healthcare utilization and infant mortality in Indonesia. The study found that contributory health insurance positively impacts outcomes, while subsidized insurance does not. The causal forest algorithm revealed geographical variations in these impacts, suggesting the need to redesign subsidized health insurance eligibility criteria [34]. Another study used causal forest analysis to evaluate the impact of health insurance programs on children’s health, finding that disadvantaged mothers benefit more from the program for preschool children [36].

In addition, Health Insurance Claims (HIC) data can be analyzed using AI to evaluate healthcare utilization, compare treatment effects, and determine patient characteristics consuming significant healthcare resources. This approach improves cost-effectiveness and decision-making in the healthcare system [11]. AI can also facilitate claims review processing and provide policy information for insurance decision-making, enhancing resource allocation efficiency and long-term health outcomes [22]. For example, one study used AI to detect ADR signals of Celecoxib compared to other analgesics, validating these signals in the Korean claims database [37].

Diagnosis and medical revolutions

AI-based models have a revolutionary role in diagnosis and patient care management. One study highlighted AI models for predicting chronic kidney disease (CKD), identifying diabetes, age, gout, and certain medications as key risk factors. The convolutional neural networks (CNN) model performed best, suggesting AI’s potential in early detection and monitoring of CKD [38]. Another study explored AI in a multilevel referral system in healthcare centers, suggesting that an expert system could improve knowledge sharing and learning, particularly for tuberculosis (TB) diagnosis [39]. Additionally, AI can predict chronic diseases and treatment expenses using data mining tools like WEKA, aiding health insurance companies in cost prediction [40].

ML methods can identify patients who benefit from traditional diagnosis methods and provide more information about drug indications to improve effectiveness. AI can streamline administrative tasks, saving time and money in diagnostics [8, 41]. AI-based models can also recommend laboratory tests based on electronic health records (EHRs), ensuring patient safety [42]. AI enhances patient care and education in health insurance by evaluating Health Insurance Claims (HIC) data. It helps prevent and detect non-communicable diseases (NCDs), i.e., cardiovascular disease, through targeted recommendations. Furthermore, AI optimizes healthcare spending by analyzing prescription patterns and assessing socioeconomic status and follow-up visits, although data security and infrastructure concerns limit the use of HIC data [11].

Hospital-based data can provide more detailed insights. For instance, AI tools are vital in gathering and analyzing data on postoperative wound infections in soft tissue sarcoma [43]. Additionally, AI can analyze data to identify treatment trends, such as the prescription patterns of Lenvatinib for advanced hepatocellular carcinoma (HCC) based on patient characteristics and policy changes [44]. AI is effective in predicting preterm birth (PTB), with maternal heart disease, age, and socioeconomic status as key factors [45]. It can also analyze gender and exercise effects on colorectal cancer prevalence, aiding in evidence-based program development [46]. Big data analysis with AI can detect disabilities early and establish preventive measures [47].

AI has been used to evaluate COVID-19 patients’ clinical progress, linking report data and health insurance claims to better understand survival rates and mortality risk factors [48]. AI in health insurance can identify sex-specific differences in pancreatic cancer risk, aiding personalized treatment and prevention [49]. For example, AI detected adverse drug reaction (ADR) signals of rosuvastatin compared to other statins using a data-mining approach based on relative risk (RR) [50].

Private insurance applications

AI in health insurance can identify subgroups within the uninsured demographic, allowing for targeted policy actions and improved health outcomes. This approach provides tailored services and products to specific groups, reducing costs and improving access to healthcare [51]. Additionally, AI can classify customers into categories based on their stability with insurance providers, using outliers and insurance dynamics to understand customer behavior [2]. Furthermore, AI can identify clusters of individuals without private health insurance, providing insights for specific policy actions, which can inform public policies to improve healthcare access for these individuals [51].

A framework for predicting potential life insurance policyholders using a data mining approach has also been introduced. AI addresses information asymmetry and improves risk selection, allowing insurers to offer personalized policies and pricing based on an individual’s health profile. This helps insurers manage portfolios better and reduce adverse selection, leading to more accurate and fair pricing for policyholders [52].

Risk management

The use of AI in health insurance enables the evaluation of large amounts of patient data, which can aid in the early detection of risk factors and assist in timely and accurate diagnoses. AI-powered healthcare applications can analyze symptoms, diagnose ailments, and potentially predict future diseases [8]. For instance, a study discussed the use of Geographical Information Systems (GIS) to evaluate risk in life insurance by collecting data on factors like air pollution, industrial areas, COVID-19, and malaria. This GIS software model aids insurance providers in making better decisions, enhancing risk analysis and decision-making in the life insurance market [53].

AI also plays a crucial role in health insurance organizations by improving patient preferences, service quality, and financial considerations [13]. A study examined factors contributing to wound infections in individuals with soft tissue sarcomas, emphasizing the importance of health insurance coverage and various factors influencing healthcare utilization, especially during the COVID-19 pandemic [43]. Another study utilized interconnected databases to address the relationship between chronic comorbidities and death in patients with COVID-19 in Korea, aiming to identify high-risk groups for preemptive interventions and provide data for recommending immunization [54]. Additionally, research analyzed big data to investigate the impact of angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin II receptor blockers (ARBs) on COVID-19 risk, finding that females, older individuals, those with low income, and recently hospitalized patients had a higher risk of infection [55].

Furthermore, AI can help in understanding the correlation between biology, lifestyle, environmental exposures, and health outcomes. For example, a study indicated that sociodemographic factors and lifestyle behaviors are associated with the incidence of symptomatic hemorrhoids, providing basic data for understanding the disease [56]. Another study discussed the challenges of collecting and maintaining high-quality data for risk analysis in health insurance, suggesting that using AI, even with limited parameters, can provide some level of risk analysis. However, identifying more risk parameters is recommended to better understand the shared risk variables that impact risk dynamics [57].

Technical analysis

AI in healthcare faces significant challenges, particularly in analyzing complex and unstructured data such as e-health records, medical imaging, text, and IoT sensor data. Traditional statistical methods require feature extraction, but deep learning, especially convolutional neural networks, offers end-to-end learning from this data, improving analysis accuracy and efficiency [58]. The increasing digitalization of healthcare data, exemplified by resources like the International Genome Sample Resource and the Cleveland Clinic’s use of data for emergency patient needs, highlights the growing importance of advanced data analytics in healthcare [59].

In the health insurance sector, AI and ML techniques, i.e., clustering methods, help model local and global relationships in data. These methods improve the analysis of health insurance coverage and identify disparities among sub-populations, enhancing decision-making processes [60]. Additionally, the adoption of wearable technology in health insurance is influenced by factors like technology policy, organizational culture, and strategic philosophy. This adoption impacts firm performance and competitive advantage, demonstrating strong relationships with most variables, except for a moderate relationship with product innovation [61]. Furthermore, AI plays a crucial role in enhancing health insurance services by proposing reforms in data processes, privacy protection, and data quality, ultimately leading to better performance and service delivery [18].

Value creation

The use of AI in health insurance is primarily focused on analyzing bills and expenditures to determine appropriate pricing for medical services, ensuring sustainable healthcare provision [12]. AI positively impacts various aspects of the insurance value chain, including pricing, underwriting, marketing, claims management, and after-sales services [31]. This role in value creation is significant, shifting from companies to consumers by providing essential information [62]. AI in health insurance utilizes behavioral analysis to identify market opportunities and target specific customers, allowing insurers to expand their customer base and offer personalized services [7, 13].

For instance, a study using data from the National Health Insurance Sharing Service Database analyzed patients with osteoarthritis (OA) to determine alternative cutoff thresholds for managing OA symptoms. This analysis, conducted using SAS software, involved receiver operating characteristic curve analysis to identify optimal thresholds [63]. The National Health Insurance Sharing Service (NHISS) in Korea provides a comprehensive database of medical histories, enabling researchers to conduct various studies and overcome previous data limitations [63]. AI’s impact on health insurance extends beyond traditional technology acceptance models, influencing customer perceptions and various elements of insurance business models, including value propositions, profit formulas, and company resources and processes [62].

Furthermore, AI has the potential to reduce healthcare expenses through advanced data management, improving financial viability and product effectiveness for insurers [64]. By utilizing healthcare big data [65], insurers can address moral hazard and adverse selection, set sustainable premiums, control costs, and enhance operational efficiency. However, the healthcare industry’s informatization level remains limited [18].

Ethical challenges and responsibilities in AI for health insurance

AI in the insurance industry offers numerous advantages but also presents significant ethical challenges. One study has underscored the ethical implications of data mining by health insurance companies, particularly concerning privacy issues [66]. To address these concerns, principles for AI governance, including trustworthiness, openness, and evidence-based risk prediction models, have been proposed [52]. Regulations are crucial to safeguarding personal health data, emphasizing the importance of privacy and data security [18]. Although AI can enhance efficiency by using public metadata to assess customers, such as analyzing social media images for claim verification, it raises ethical concerns that need to be addressed [7].

Ensuring data security and confidentiality is vital when integrating AI into health insurance. Studies have highlighted risks such as cyber-attacks and adversarial attacks, emphasizing the need to address these vulnerabilities [31, 64]. Transparency in AI decision-making processes is crucial, and tools that provide consumers with insights into their life-scores can enhance this transparency [32].

Insurance companies leveraging AI need to be held accountable for protecting customer identities against cybercrime [64, 67]. Therefore, establishing an ethical and regulatory framework for the safe application of big data analytics and AI in health insurance is imperative, while clear data governance policies, legal standards, and a human-centered approach are necessary. Collaboration between insurers, governance bodies, regulators, and policymakers ensure transparency and accuracy in the development of big data analytics. Without a sound ethical environment, the use of such analytics could result in detrimental consequences [68].

This study, while methodologically robust, has some limitations. The search was limited to three databases (PubMed, Web of Science, and Scopus), which may have excluded relevant studies from other sources. Additionally, focusing solely on English-language publications might have narrowed the scope and excluded insights from diverse regions.

While ethical challenges such as data privacy, algorithmic bias, and equitable access were highlighted, exploring these concerns comprehensively was beyond the scope of this study. Future research should investigate these issues further and examine governance frameworks to ensure responsible AI integration into health insurance systems.

Conclusion

The integration of AI into health insurance has driven significant advancements across critical areas, including policy design, cost management, fraud detection, and operational efficiency, as demonstrated in this study’s findings. AI tools provide deeper insights into health insurance programs, enabling personalized services, healthier lifestyles, and optimized financial operations through cost reduction and enhanced pricing strategies. Additionally, AI strengthens transparency, data security, and customer privacy, supporting fairer practices and reduced operational costs.

However, the implementation of AI is not without challenges. Ethical concerns such as data privacy, algorithmic bias, and equitable access require robust regulatory frameworks to ensure trust and accountability. Safeguarding personal health information and addressing ethical dilemmas are essential for fostering public confidence in AI-powered systems.

Future research should focus on addressing these ethical challenges comprehensively while exploring consistent definitions and methodologies for AI applications in health insurance. Practical implications include fostering collaboration among insurers, regulators, and policymakers to develop actionable governance strategies and sustainable integration models. By tackling these issues, the health insurance sector can fully leverage AI’s potential to revolutionize operations, improve accessibility, and deliver better customer experiences.

Abbreviations

AI

Artificial intelligence

IoT

Internet of things

ML

Machine learning

Author contributions

AT and MR-a conceived the study. AT supervised all evaluation phases and critically revised the manuscript; he is the guarantor. MR-a wrote the main manuscript text and prepared figures. AT, MM, RD, and HRR provided feedback on the result and edited the manuscript. MR-a and AB categorized AI applications independently and created descriptions by synthesizing the extracted information. RD, HRR, AB, AAF, MM, MR-b, AO, SS, SB, and HM edited the manuscript. All authors reviewed the manuscript.

Funding

Tehran University of Medical Sciences (TUMS) funded this research.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Ethical Committee of the Tehran University of Medical Sciences; all methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.