IMF researchers propose to have AI use ONLINE history to determine credit rating, in name of ‘inclusion’.

Sharing is Caring!
Traditional data used for credit scoring could be replaced by AI-powered processes based on online search and purchase histories to better reach a billion-plus “unbanked people” globally, a group of IMF researchers has proposed.

Using non-financial data such as “the type of browser and hardware used to access the internet,” or “the history of online searches and purchases,” which is then fed into technology powered by AI and machine learning, could “advance financial inclusion, by, for example, enabling more credit to informal workers and households and firms in rural areas,” write Arnoud Boot, Peter Hoffmann, Luc Laeven and Lev Ratnovski in a post on the International Monetary Fund (IMF) blog.

The authors argue these “alternative data sources are often superior than traditional credit assessment methods,” which tend to boost credit expansion during booms and shrink it during busts. Moreover, they say, traditional credit data is not available for “certain kinds of people, like new entrepreneurs, innovators and many informal workers,” and even well-off immigrants to the US.

Their AI-powered credit score based on web searches is just one of the proposals in the paper, which argues the Covid-19 pandemic is “turbo-charging” the technological transformation of the financial industry and looks at how the IMF can “get ahead” of the problems that might create.

Boot is a professor of finance at the University of Amsterdam in the Netherlands. Hoffman and Laeven are economists with the European Central Bank, where Ratnovski has been seconded from his job as an IMF economist. In other words, these are very serious finance people and their proposal isn’t meant in jest. They just don’t seem to be familiar with the “when I die, delete my internet history” meme.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.