Discrimination and AI in insurance: what do people find fair? Results from a survey

• URL: http://arxiv.org/abs/2501.12897v1
• Date: Wed, 22 Jan 2025 14:18:47 GMT
• Title: Discrimination and AI in insurance: what do people find fair? Results from a survey
• Authors: Frederik Zuiderveen Borgesius, Marvin van Bekkum, Iris van Ooijen, Gabi Schaap, Maaike Harbers, Tjerk Timan,
• Abstract summary: Two modern trends in insurance are data-intensive underwriting and behavior-based insurance. Survey respondents find almost all modern insurance practices that we described unfair. We reflect on the policy implications of the findings.
• Score: 0.0
• License:
• Abstract: Two modern trends in insurance are data-intensive underwriting and behavior-based insurance. Data-intensive underwriting means that insurers use and analyze more data for estimating the chance that a consumer files a claim and calculating the premium based on that estimation. Insurers analyze the new datasets with artificial intelligence (AI) to discover new correlations, with which they can estimate the policyholder's expected claims cost more precisely. Insurers also offer behavior-based insurance. For example, some car insurers use AI to follow the driving behavior of an individual policyholder in real-time and decide whether to offer that policyholder a discount. Similarly, a life insurer could track a policyholder's activity with a smart watch and offer a discount for an active lifestyle. In this paper, we report on a survey of the Dutch population (N=999) in which we asked people's opinions about examples of data-intensive underwriting and behavior-based insurance. The main results include the following. First, if survey respondents find an insurance practice unfair, they also find the practice unacceptable. Second, respondents find almost all modern insurance practices that we described unfair. Third, respondents find practices fairer if they can influence the premium. For example, respondents find behavior-based car insurance with a car tracker relatively fair. Fourth, if respondents do not see the logic of using a certain consumer characteristic, then respondents find it unfair if an insurer calculates the premium based on the characteristic. Fifth, respondents find it unfair if an insurer offers an insurance product only to a specific group, such as car insurance specifically for family doctors. Sixth, respondents find it unfair if an insurance practice leads to higher prices for poorer people. We reflect on the policy implications of the findings.

Related papers

• Classification problem in liability insurance using machine learning models: a comparative study [0.0]
  We apply several machine learning models to classify liability insurance policies into two groups: 1 - policies with claims and 2 - policies without claims. In this study, we apply several machine learning models such as nearest neighbour and logistic regression to the Actuarial Challenge dataset used by Qazvini.
  arXiv  Detail & Related papers  (2024-11-01T04:35:39Z)
• A Bayesian Approach for Prioritising Driving Behaviour Investigations in Telematic Auto Insurance Policies [0.6249768559720121]
  We make use of trip GPS and accelerometer data, augmented by geospatial information, to train an imperfect classifier for delivery driving on a per-trip basis. A posterior probability is converted to a priority score, which was used to select the most valuable candidates for manual investigation. The approach has achieved a significant improvement in efficiency of human resource allocation compared to manual searching.
  arXiv  Detail & Related papers  (2024-04-22T15:26:24Z)
• Harnessing GPT-4V(ision) for Insurance: A Preliminary Exploration [51.36387171207314]
  Insurance involves a wide variety of data forms in its operational processes, including text, images, and videos. GPT-4V exhibits remarkable abilities in insurance-related tasks, demonstrating a robust understanding of multimodal content. However, GPT-4V struggles with detailed risk rating and loss assessment, suffers from hallucination in image understanding, and shows variable support for different languages.
  arXiv  Detail & Related papers  (2024-04-15T11:45:30Z)
• AI, insurance, discrimination and unfair differentiation. An overview and research agenda [0.6144680854063939]
  Insurers seem captivated by two trends enabled by Artificial Intelligence (AI) Insurers could use AI for analysing more and new types of data to assess risks more precisely: data-intensive underwriting. Insurers could also use AI to monitor the behaviour of individual consumers in real-time: behaviour-based insurance. While the two trends bring many advantages, they may also have discriminatory effects on society.
  arXiv  Detail & Related papers  (2024-01-22T12:39:36Z)
• Measuring and Mitigating Biases in Motor Insurance Pricing [1.2289361708127877]
  The non-life insurance sector operates within a highly competitive and tightly regulated framework. Age-based premium fairness is also mandated in certain insurance domains. In certain insurance domains, variables such as the presence of serious illnesses or disabilities are emerging as new dimensions for evaluating fairness.
  arXiv  Detail & Related papers  (2023-11-20T16:34:48Z)
• A Counterfactual Safety Margin Perspective on the Scoring of Autonomous Vehicles' Riskiness [52.27309191283943]
  This paper presents a data-driven framework for assessing the risk of different AVs' behaviors. We propose the notion of counterfactual safety margin, which represents the minimum deviation from nominal behavior that could cause a collision.
  arXiv  Detail & Related papers  (2023-08-02T09:48:08Z)
• Incentive-Theoretic Bayesian Inference for Collaborative Science [59.15962177829337]
  We study hypothesis testing when there is an agent with a private prior about an unknown parameter. We show how the principal can conduct statistical inference that leverages the information that is revealed by an agent's strategic behavior.
  arXiv  Detail & Related papers  (2023-07-07T17:59:01Z)
• Evaluating COVID-19 vaccine allocation policies using Bayesian $m$-top exploration [53.122045119395594]
  We present a novel technique for evaluating vaccine allocation strategies using a multi-armed bandit framework. $m$-top exploration allows the algorithm to learn $m$ policies for which it expects the highest utility. We consider the Belgian COVID-19 epidemic using the individual-based model STRIDE, where we learn a set of vaccination policies.
  arXiv  Detail & Related papers  (2023-01-30T12:22:30Z)
• A Data Science Approach to Risk Assessment for Automobile Insurance Policies [1.0660480034605242]
  We focus on risk assessment using a Data Science approach. We predict the total claims that will be made by a new customer using historical data of current and past policies.
  arXiv  Detail & Related papers  (2022-09-06T18:32:27Z)
• Risk Minimization from Adaptively Collected Data: Guarantees for Supervised and Policy Learning [57.88785630755165]
  Empirical risk minimization (ERM) is the workhorse of machine learning, but its model-agnostic guarantees can fail when we use adaptively collected data. We study a generic importance sampling weighted ERM algorithm for using adaptively collected data to minimize the average of a loss function over a hypothesis class. For policy learning, we provide rate-optimal regret guarantees that close an open gap in the existing literature whenever exploration decays to zero.
  arXiv  Detail & Related papers  (2021-06-03T09:50:13Z)
• Post-Contextual-Bandit Inference [57.88785630755165]
  Contextual bandit algorithms are increasingly replacing non-adaptive A/B tests in e-commerce, healthcare, and policymaking. They can both improve outcomes for study participants and increase the chance of identifying good or even best policies. To support credible inference on novel interventions at the end of the study, we still want to construct valid confidence intervals on average treatment effects, subgroup effects, or value of new policies.
  arXiv  Detail & Related papers  (2021-06-01T12:01:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.