Global menu

Our global pages

Close

6th Annual Digital Financial Services and Fintech Conference: Artificial Intelligence

  • Global
  • Financial services - Digital Financial Services

13-01-2021

 

Simon Gamlin moderated a panel discussion with Charlotte Walker-Osborn, Simon Walker and Orlando Machado. Together, they challenged the common ‘all powerful robot’ stereotype that springs to mind when we hear the word ‘AI’, pointing out that in reality, the term often refers to gradually more sophisticated versions of the ‘machine learning’ processes that have existed within financial services firms for some time.

The panel also highlighted that ‘explainability’ is key to breaking down the barriers in respect of the use of AI.  Orlando also questioned the wide use of AI jargon such as ‘Big Data’ with these generic terms often being unhelpful in the industry for people to understand what is happening in practice.  Simon W noted his experience of firms that Kubrick work with more frequently calling for a ‘data literacy standard’; one that can be measured as a metric across both senior managers and the wider employee base.  They discussed that such upskilling in data literacy could also help combat the unrealistic apportioning of risk in contracts (such as firms requesting wide-ranging unlimited liability positions).  Charlotte further noted that explainability is at the heart of effective regulation – in order to achieve a regulatory landscape that reflects actual use of AI, lawmakers must therefore engage with those on the ground that are using and developing the technology.

Finally, the panel discussed how AI can assist workplaces in becoming more diverse, flagging that firms must be mindful of the range of data they input into their AI systems; particularly in the context of attracting Gen X/millennials as part of job application processes.  Making conscious data choices will allow firms to harness a more creative, diverse workforce and in turn also tap into the behaviour of an equally broad customer base – whilst a failure to do so could result in unwanted bias and skewed outcomes.