Financial models relying on Artificial Intelligence (AI), particularly its Machine Learning (ML) branch, are in the high risk-complexity-materiality spectrum in the model inventory of banks and insurers. Some of these have increased their popularity with the advent of Big Data analytics.

To establish common ground, ML applies and refines – or trains – a series of algorithms on a large dataset by optimizing iteratively as it learns in order to identify patterns and make predictions for new data1.

Financial services and other industries are expected to increase the use of such Big Data from 7.9 ZB in 2016 to 44 ZB by 20202. They are also expected to increase the use of non-traditional data and change the mix of internal and external data used.

Such use of data and reliance on the results they produce – including diverse commercial uses that involve analytics, model benchmarking and decision-making – is likely to increase with ~55% of the insurers using or experimenting with AI and ML solutions5.

Consequently, attention is focussed on how financial institutions and regulators are responding in terms of governance of these models and providing appropriate oversight while maintaining proportionality to conduct model development and validation activities.