Insurers possess, and have access to, vast historical – and ever-growing – datasets, which make them prime candidates for the adoption of artificial intelligence (AI)1 and machine learning (ML)2 for risk modelling.

That potentiality, however, did not manifest until the Covid-19 pandemic pushed insurers to scale up data mining. This trend is only expected to continue, and accelerate. 

While global demand for AI/ML is expected to increase ~40% annually through this decade3 , there is a proxy signal for incipient insurer-specific interest: searches for the term ‘Insurance AI’ on Google Trends have shot up 400% over the past five years. 

Practitioners are looking at other informal ways to learn; sites like Kaggle have 2200+ AI/ML competitions and thousands of data sets available for different use cases around data science (DS) and guidance-notebooks for Python, R and Julia languages, as of August 2021. Learning sites like LinkedIn have increased their AI/ML course offerings to 110+ compared with 30-50% less courses 18 months ago when the pandemic started, clearly showing the ongoing interest in this area. Other learning sites like Udemy, Google, Coursera, Microsoft EdX and Udacity too, have increased similar course offerings.