What Has AI Done for Financial Services?
Everyone's talking about the potential for artificial intelligence in the enterprise, but what's actually been achieved to date? Are the majority of companies comfortable with the technology and how it can be used?
Dermot O’Kelly, Senior Vice President, Europe at Finastra
Think your organization hasn’t embraced AI? Think again. The reality is that there are hundreds of applications of artificial intelligence embedded in everyday organizational life. From pay-per-click ads to social listening, chatbots to lead scoring, biometric security to network attack detection. As Europe at Finastra's Senior Vice President Dermot O'Kelley outlines below, the chances are that your organization is already relying heavily on AI for a range of functions.
It’s true that many of these services may be provided by third parties connecting directly to systems via open APIs. The organization therefore doesn’t need to become the expert. In fact, there is a proliferation of external experts as AI becomes ever more accessible. In less than two years, training time for machine vision algorithms dropped by over 99%. It went from three hours to just 88 seconds – whilst computational costs dropped from ‘thousands of dollars to double-digit figures’.
It therefore comes as no surprise that organizations are looking at how they can benefit from the AI revolution, to help boost areas such as operational efficiency, security, predictive capabilities, product development or customer satisfaction.
In less than two years, training time for machine vision algorithms dropped by over 99%.
Leading the way is the financial services sector, not least because of the vast amounts of data held by legacy organizations, but also in response to the changing expectation of consumers. Tech giants created new models of engagement, platforms that consolidated services and captured data to further fuel predictive capabilities, and this expectation of convenience is now shifting to financial services, where consumers are now more than comfortable with concepts like robo-advisory. Institutions, regardless of whether they’re providing retail services, lending, trade finance, wealth or any other line of business, are racing to adopt similar models without relinquishing customer data.
As data proprietors, the world of opportunity that AI affords any organization is immense. Data is the new currency as we enter the fourth industrial revolution, and all AI applications rely on huge amounts of data to function well. So, why aren’t all organizations rushing to embrace AI?
- Firstly, the lack of appropriate, compliant data. Successful AI models require tens of thousands of records, cleaned and labelled, that have been captured and stored according to strict regulatory conditions.
- Secondly, the growing urgency for transparency and accountability in data, models and decisions to stop the cumulative effect of bias. Best practices are offered, but as yet there is no universal regulation or consensus.
- Thirdly, the ethical implications - just because you can doesn’t automatically mean that you should, particularly where privacy is concerned. As Cap Gemini noted, consumers are attuned to ethical practices…get it wrong and the trust – and business – is lost.
- Finally, the well-documented shortage of skilled AI and Data experts. Thankfully, through open APIs, all organizations can connect in AI-based capabilities developed by specialist third parties – it doesn’t always pay to build in-house.
[ymal]
The intelligence race continues unabated, with escalating VC investment in AI and new, exciting applications that are having tangible success. Still not sure what Artificial Intelligence can do? Very soon it will be easier to recall the few things the technology can’t do.
