The Fairness of Machine Learning in Insurance: New Rags for an Old Man?

  title={The Fairness of Machine Learning in Insurance: New Rags for an Old Man?},
  author={Laurence Barry and Arthur Charpentier},
Since the beginning of their history, insurers have been known to use data to classify and price risks. As such, they were confronted early on with the problem of fairness and discrimination associated with data. This issue is becoming increasingly important with access to more granular and behavioural data, and is evolving to reflect current technologies and societal concerns. By looking into earlier debates on discrimination, we show that some algorithmic biases are a renewed version of older… 

Tables from this paper


Fairness in Uncertainty: Some Limits and Misinterpretations of Actuarial Fairness
The recent proliferation of new data and technologies enables increasingly finer personalization of products and prices in every domain. In insurance, this revives and enlarges old debates around
Risk Classification's Big Data (R)evolution
While big data may be a natural next step in risk classification, it may require a revolutionary approach to regulation and regulators will have to increase their capacity to analyze the data inputs, algorithms, and outputs of the classification schemes.
Inherent Trade-Offs in the Fair Determination of Risk Scores
Some of the ways in which key notions of fairness are incompatible with each other are suggested, and hence a framework for thinking about the trade-offs between them is provided.
Rate Making and Society’s Sense of Fairness
Premium differentiation on the basis of observable risk factors has always been one of the main activities of the rate making actuary. Conflicts may arise between the differentiation proposed by the
Enacting Actuarial Fairness in Insurance: From Fair Discrimination to Behaviour-based Fairness
ABSTRACT In line with developments in the personalisation of risk, the idea that insurance products should above all be ‘fair’ to the policyholders is increasingly voiced by commentators. The
Fairness and Equality in Insurance Classification
Insurance is to a large extent based on risk selection and classification. Legislators however are inclined to impose restrictions to these differentiations by banning those that are considered to be
Fairness in Machine Learning: Lessons from Political Philosophy
This paper draws on existing work in moral and political philosophy in order to elucidate emerging debates about fair machine learning.
The Moral Limits of Predictive Practices: The Case of Credit-Based Insurance Scores
Corporations gather massive amounts of personal data to predict how individuals will behave so that they can profitably price goods and allocate resources. This article investigates the moral
Disparate Impact and Unfairly Discriminatory Insurance Rates
Historic actuarial literature, general insurance literature, and legislative histories reveal “unfairly discriminatory rates” to be a cost-based concept. A rate structure is unfairly discriminatory
The Shifting Rhetoric of Insurance Denial
The focus on actuarial tables and rating systems by state regulators is insufficient to ensure that protected groups are not discriminated against by the insurance industry. The most powerful tool