With any new technology, there are arguments in favor of being both a leader and a fast follower.
In the case of Australia's financial sector, they have no choice. Providers in the U.K. and the U.S. have gained such a lead on them that they have almost no choice than to be fast followers.
A bank like the Royal Bank of Scotland, for example, is being scrutinized very carefully in the antipodes because it is an acknowledged leader in AI implementation. The bank’s digital teller, named “Cora,” has ear piercings and varied facial expressions and can answer up to 200 direct queries from customers on laptops, tablets and smartphones.
Understanding Where AI Matters
In Australia, the banking pilots are a little less ambitious as the banks dip their toes into the waters of AI. None of the banks are committing significant resources as yet. Instead, they are choosing individual projects and running proof of concept pilots before they make their next move.
Partly, that is driven by fear of failure, and of over-committing on budgets which are already under significant pressure. There are also a lot of questions to answer before banks make a move.
Should they act now or wait until there is something potentially better around the corner? Do they have a sufficient understanding of the new business risks which will be introduced? Are they solving a client problem or just automating a lousy process which needs re-engineering?
Perhaps more importantly, if they do nothing will they be left behind?
In the case of call centers, it does make sense to start with a finite project. Many of the queries which come into call centers are the same, and there are probably 10 main conversations customers want to have.
It makes sense then to teach the AI how to respond to these queries because the more conversations the AI hears, the better it gets at answering them. It is sensible that the call center bots become expert in answering the 10 most common queries and leave the exceptions to humans who can add value.
Once the AI masters those 10 basic questions and proved its worth, it justifies moving on to other queries and processes.
They might be behind the global curve, but some of the Australian asset managers linked to the major banks have moved in this direction, even though a chronic lack of skilled people hampers them.
At one major call center, AI has reduced the time to retrieve information on customer pension accounts from about 30 minutes by manual processes down to three minutes.
Advisers call in to ask call center agents about their customers' pension accounts. Where it used to take the agent time to collate all that information from across multiple systems, the AI can retrieve it faster.
It is also proven to be more accurate once retrieved. Instead of the agent jotting the details down on a pad and then reading them over the phone or writing them in an email, the AI creates a PDF that is sent to the adviser and which goes into the customer file.
Another major push for AI in Australian financial services is regulatory, spurred on by a Royal Commission into misconduct in the banking industry which has unearthed multiple examples of bad bank behavior.
AI and Robotic Process Automation (RPA) are being pushed hard as part of the solution to many of these problems.
Regulators like AI because it is 100 percent auditable, and providers like it because with one keystroke they can suck data out of multiple systems.
AI is also being trialed in fraud detection, and with some success. Machines are much better and faster than picking up anomalies across multiple systems, and this business case is crying out for AI.
The caveat here, of course, is in the quality of the data in that the banks have some way yet to go.
Offshore to Onshore
Looked at in overview, Australia’s banks are global laggards with AI, but in domestic terms they are leaders.
It is a long play, but the trend has significant implications for offshore Business Process Outsourcing (BPO). While some companies are using BPO providers to help automate processes, others are doing that and then building in-house Centers of Excellence that could ultimately undermine the BPO relationship.
As one digital executive put it, AI and RPA are bringing processes back onshore after years of sending them offshore.