AI - For the better?

By Teamspirit on Monday, 4 February 2019

__AI is starting to transform the financial services landscape, offering as it does the ability to reduce cost, improve accuracy, prevent fraud and cut the need for that most expensive of resources, actual human beings. There’s no denying the appeal of something that can shorten claim times, drive down the costs of transactions and make the process of applying and buying, easier and more intuitive.

But AI also raises the prospect of financial service providers having the ability to cherry-pick and choose to serve only those markets and segments of markets that are most profitable and least risky for them. It’s already being used for risk analysis and to identify and stop fraudulent claims in insurance; but concerns have been raised that it is being used to improve margins by pricing out those it ‘learns’ are more likely to claim – often on the flimsiest of indicators such as a surname.

The combination of capability and autonomy it delivers, means that AI is making choices with profound consequences for customers and employees alike. Transparency is now at risk, when the complexity of the algorithms sometimes means that the answer is literally ‘because the computer says no’.

In 2019, businesses must think carefully when applying AI to their processes. Each will have to balance their need to protect margin and improve profitability, with the need to act responsibly and be able to explain how decisions are made.


Implications

• AI will continue to slash acquisition costs throughout financial services, while its predictive analytics will provide greater security.

• The complexity of the algorithms means that transparency is at risk. Companies will need to clearly evidence when introducing AI, that they are acting responsibly and for the overall good.

• Without careful consideration, AI could lead to serious societal exclusion and customer detriment issues.

Related News

Thu 21 Mar 2024

In marketing, are we losing sight of ‘Business to Human?

Read more

Fri 9 Feb 2024

Going social for advice. Should advisers respond… or join the party

Read more

Tue 2 Jan 2024

2024 Signals Report

Read more

Tue 14 Nov 2023

Branding with AI: The good, the bad and the reality

Read more