The recent decision by a Nordic firm to stop using a chatbot for client onboarding doesn’t mean that AI tools will not dramatically change financial services, a practitioner argues.

Science fact appears to have caught up with science fiction with the report a few months ago that online bank Nordnet had “fired” its chatbot known as Amelia for not performing her job well enough. 

While the story (source: Nordic Business Insider, 9 July, 2018) might have provoked a few chuckles from technophobe humans, it is a cautionary tale about the limits of what AI can do, a wealth management practitioner says. 

The media is abuzz with stories about how artificial intelligence, machine learning and robots are going to make bankers redundant. The power of tech to augment human capabilities is real enough, but perspective is necessary, argues Paul Ferreira, senior ICT architect at Maitland. (Maitland is the global advisory and funds administration firm which recently launched a family office organisation serving clients in South Africa.)

Ferreira is responsible for automating and integrating technology ideas in the fund administration side of Maitland’s business.

Asked about the limits and potential of AI, Ferreira said the Nordnet example suggested that some clients are not happy to be served by algorithmically-powered chatbots and other platforms. “Banks are paying a very expensive price because some people prefer the human touch,” he told this publication in a call. 

Another problem is automation involving DIY form filling online for onboarding; this produces a lot of client pushback, he said. “We’ve seen people trying to automate that process…but they have been caught out because they are not interested enough in training the robots,” he continued. “Robots are `digital employees’ and they require some management”.

The idea that HR managers will need to learn how a robot works with a human sounds bizarre at first, but these digital assistants need to be integrated into a team of humans. Ultimately, information processed by a machine is the responsibility of the humans who run and own financial institutions, creating a new set of potentially tricky legal responsibility questions, rather as self-driving cars might do as and when accidents occur. (This news service produced research earlier this year about the impact of AI on wealth management.)

Ferreira said that Maitland has a system with the name of “Eric”, taking its name from the firm’s founder, Eric Pfaff. 

“We have employed Eric internally and that’s where we have got the gains in terms of our profitability,” Ferreira said. “Using a graduate just to tap into a keyboard for some data is a waste of time and that person’s intelligence.”

As far as artificial intelligence is concerned, the key is to remember that AI is only as good as the data that is fed into it, Ferreira argued. AI can be used to track and monitor clients’ behaviours and hence figure out better ways to serve them with what they want.

Even without chatbots, a number of firms are using technology in different ways to ease the pain of client onboarding. This publication has heard it said that on average, it can take up to two months to take on a new client because a great deal of compliance and related checks are required. And storing and tracking client data is a laborious task of a sort that robots can be tasked to perform. Several studies point to how new technology could branch out, such as here.

You’re fired!
Even by the standards of such stories, Nordnet’s removal of its “Amelia” co-worker was an unusual one. The character was launched in 2017, with the goal of speeding up client onboarding. However, it appeared that clients were not enthusiastic, although the character is used at a mass of companies. SEB, the Swedish bank, has used the fair-haired looking character.

The software behind Amelia is IPSoft. On its website, IPsoft refers to a study by consultants McKinsey in which it says AI technologies are expected to boost labour productivity by up to 40 per cent.

Blockchain
WealthBriefing turned the conversation to the evolving field of blockchain and other distributed ledger technologies, asking what their impact will be. Ferrerira said the impact is going to be large for several years. 

“Private wealth managers will need to prepare themselves to invest in this [blockchain] technology,” he said. One effect of such technologies is that it is putting privacy back into the conversation, and rightly so, he said.

“The sovereignty of the individual is ultimately what this is all about,” he added.