Close up image of a computer server unit with seven ports and the first and second lit up in green

Artificial intelligence (AI) is an emerging trend that has long been forecast and might being having an impact on the finance function.

AI in finance

AI itself is an encompassing term that embraces a number of technological advances including:

  • Machine learning - using neural networks, statistics and operational research to identify insights in data without being programmed what to conclude
  • Deep learning - using many layers of computing power and improved training techniques to identify patterns in data.

In delivering an AI solution you may be combining a number of technologies. For example, enhancing analytics and forecasts by using structured and unstructured data to deliver forward thinking insights. Cloud based storage may be an asset for the data volumes involved and the computational power needed.  

Organisations have examples of using forms of AI to address some of the data validation errors encountered in RPA processes by applying machine learning to the errors. Some organisations are creating process chains where data is captured through chatbots, entered using RPA tools and errors resolved using machine learning.

Implementing AI in finance

There are opportunities to apply AI in finance. For many organisations this is the level of opportunity.  One which may yet only be emerging as a proof of concept rather than as a robust solution.Perhaps the most significant issue for finance is the skills needed to support this next technological wave.

AI needs individuals who understand the data and the processing capability and how to frame the problem. This takes the skills needed in finance to another level. Bias is an issue which these individuals need to be able to recognise and address. Bias in output occurs when the input data set is not representative of the outcome and the machine 'learns' in the wrong way. This needs to be identified and corrected before reliance is placed upon the solution.

Impacts and issues

The following impacts and issues need to be considered:

  • Do you have opportunities where the use of AI could improve your decision making processes?
  • Do you have the skills within the team, or can acquire them, to make AI feasible?
  • Does your IT strategy support the computational resources needed to deliver on AI?

About ACCA lead author, Clive Webb

Further reading