PwC predicts that by 2020, consumer intelligence will be the most important predictor of revenue growth, the public cloud will be the most dominant infrastructure model, and - crucially - that Asia will emerge as a key centre of technology-driven innovation.
A Microsoft-commissioned IDC survey bears the claim out: business leaders across the Asia-Pacific (APAC) region expect 85% of jobs in their organisations to be transformed, particularly by Artificial Intelligence (AI). The Taiwanese government is investing $540 million in AI research to help its workforce remain competitive, in particular by preserving the collective knowledge of retiring employees through AI systems.
Microsoft and IDC predict that AI will be a major driver of economic growth, increased productivity, cost reduction and customer advocacy between now and 2021, supporting 40% of digital transformation initiatives. However, the same report suggests that only 7% of companies in the APAC region are positioned to lead these transformations. These businesses are concerned about how AI and other technologies will disrupt their competitiveness, they are creating cultures of innovation within their organisations, and they are aware of the challenges that lie between them and a return on investment in AI.
- You may like: Debt and debt recovery in APAC: What the politicians say
But how many of those leaders will you find within the APAC region’s banking sector? Are APAC banks ready for AI? Can it help with the particular challenges the region is facing? And are banks, debt collectors and other financial institutions ready to embrace AI?
AI in APAC
The APAC region is already a hotbed for financial technology innovation, often targeted directly at areas where daily operations slow down around human interaction. For instance: opening a bank account in Hong Kong is now automatic, thanks to a direct interface between digital ID systems and virtual banks, streamlining a process that was previously slow, paper-based and burdened with a need for hard copy documentation.
In China, Xiaoi - a major league chatbot with 500 million users and 100 billion conversations to date - has seen widespread adoption across the financial, e-commerce and telecoms industries. Xiaoi handles all manner of routine interactions that don’t necessarily need a person to make decisions; straightforward transactions like money transfers and repayments, along with data collection and user analysis by the banks. Xiaoi also provides China Merchant Bank’s knowledge base system, distributing expertise and knowledge across departments on demand.
In lending, Chinese financier Ping An uses AI ‘computer vision’ to screen loan customers for reliability. Prospective borrowers answer questions about their income and plans for repayment over a video call: the AI views those calls live, monitoring 50 tiny facial tics to determine whether they are telling the truth. By pinpointing likely defaulters, Ping An hopes to head off debt collection issues before they arise - but what can AI do when an account is already failing to perform?
AI and debt collection
The kind of screening that Ping An uses AI to carry out would be ideal for addressing the loose credit control situation in China, India and Thailand. Personal insolvency is implicated in 24% of complex debt collection cases across the APAC region: AI can help identify would-be borrowers who are insecure or less than truthful about their financial circumstances.
Such insights can also be used to shape and guide approaches to dealing with debtors. Machine learning allows banks to preserve their relationships with customers in debt: AI can quickly identify and recommend the time of day, the tone of voice, and the specific way to pay that a given customer profile is most likely to respond well to. Personalising the debt collection service like this ensures that the actual labour hours of collectors are put to the best possible use.
There’s also a question of operational efficiency to consider. As more and more banking systems become integrated with AI, the bank’s operations as a whole can become more joined up. The easier and more responsive a system is to use, the more people will use it; and with complexity being such a serious cause of loan delinquency in APC, that simplicity is sure to pay off
The reality: what can AI do for banking as a whole?
Here and now, AI is what ‘big data’ was three years ago. Industries are excited about it, innovators are obsessed with it, but very few institutions are clear about what exactly they mean by it or what they can use it for.
In many cases, banks are ahead of the game: they have worked with business intelligence tools for quite some time, and there’s less room for innovation here than outsiders generally realise.
What banks haven’t done is to automate - outside innovation hotbeds like Hong Kong, banking operations often rely on Excel, paper, and considerable labour hours. This is the application of AI that makes sense for banks - automated processing of information, assembling it and comparing it to the criteria that determine who is a low or high risk borrower. This is not a massive change to operations, but it is a significant improvement in their efficiency, with computers taking over the donkey work of routine credit checks, transaction handling and frequently asked questions.
AI can't solve all of a bank’s problems, but it can improve the efficiency with which you identify and understand the problems and deploy your solution.