Artificial Intelligence issued loans to rise $17 bn by 2021 says Juniper

A new study by Juniper Research titled AI & Machine Learning: Fintech Dynamics, Disruption & Future Opportunities 2016-2021 predicts that unsecured consumer loans that use artificial intelligence or machine learning technology will jump to 960 per cent over the next five years (2016-2021) .

Fueling this growth will be advances in analytics and accessible computing power.

The study also predicts that machine learning spend in Fintech will advance rapidly, owing to the highly data‑driven nature of the market, meaning that AI integration is likely to spell substantial benefits. Juniper researchers claim that machine learning has seen a tremendous leap in activity since 2011, with substantial increases in VC and R&D investment. For example, 2 Fintech start-ups, Kabbage and ZestFinance, have collectively raised US$500 million in funding alone. Meanwhile vendors analyzed in Juniper’s research have spent a total of US$83 billion in R&D during 2015. Each of these vendors names AI as a part of core strategy.

Juniper identified two barriers to innovation that have come down in recent years: high cost of artificial technology and access to extensive data sets for algorithm training were limited. Presently, the ability to use GPU (graphics processing unit) hardware for processing massive and highly available data sets, along with unlimited affordable computing power in the form of distributed architecture has opened the market to a swathe of disruptive new
players.

AI is particularly useful for risk-assessment purposes, where variables from numerous financial and non-financial data points are assessed by algorithms to approve loans. This widens the addressable market for financial institutions considerably over traditional FICO credit scoring, where lack of credit history may mean loan rejection despite a real low risk for the lender.

“Where Big Data analytics offered retrospective business intelligence, machine learning offers predictive and even prescriptive capabilities,” noted research author Steffen Sorrell. “Data is key – and industries able to draw expertise from data scientists will be the first to capitalize on the AI opportunity.”

Juniper Research says AI has a long history in financial institutions. As far back as 1980 expert systems were around to help perform ‘if this then that’ algorithms to data to support decision-making. Today, machines are being used to detect potentially fraudulent card transactions.