Banking and Risk - Artificial Intelligence is nothing new

Artificial Intelligence (AI) is a discipline that has been revived. In the 1980s Artificial Intelligence and Expert systems were very fashionable and the British Computer Society had an Expert Systems Specialist Group run by the famous Alexander D'Agapeyeff. Other Scientists with great names doing pioneering work at this time included Robert Kowalski and Edward Feigenbaum. Everyone was talking about the fifth generation and AI was going to save the planet! (it still might). The management consultancy where I worked at the time was looking for opportunities to use AI in a practical way in banks.

A certain Central Bank wanted a tool that would let them know when a financial institution was at serious risk of going bust. A system was selected which the inspectors could use to build an advanced risk model into which data from regulatory reports could be fed. Factors covered included solvency, liquidity, asset / liability types, maturities, currency, interest rate and counterparty exposures and more. The knowledge and wisdom of the inspectors was defined in logic rules using an expert system shell. The result was the introduction of an Artificial Intelligence based expert system that could be used as an early warning solution for risk capital management in banks and financial institutions. This was far ahead of its time – a forerunner of JP Morgan risk metrics and others - and we are talking about 1987!