Predictive Model
A model that estimates a future outcome or likelihood from historical data, such as the probability of a claim or the risk of a policyholder.
A predictive model is a model that uses historical data to estimate future outcomes or probabilities. In insurance, predictive models are used to estimate the likelihood of a claim, the expected cost of a policy, the risk of fraud, or the probability of policyholder attrition.
Predictive models are not new to insurance. Actuarial models have long been used for pricing and reserving. What is new is the scale of data, the complexity of machine-learning techniques, and the use of external consumer data that may not be fully understood.
Regulators treat predictive models as AI systems when they influence consumer decisions. That means they must be inventoried, validated, tested for fairness, and monitored. See our glossary entries on underwriting and pricing, model validation, and training data.