The Age of AI is finally upon us. Arguably the single biggest technology revolution of the 21st century, artificial intelligence (AI) is now coming to full maturity with many companies and organisations racing fast to harness its full potential. And the financial industry is currently leading the charge.
This technology, which enables machines to provide human-like intelligence on a quantum level in order to augment human intelligence and decision making, is now being used to cut costs, improve compliance, increase revenue and improve customer experience.
Its business value is so huge that anyone who fails to put a solid strategy behind it is expected to be edged out by competition. In fact, the business value that could be gained through the use of artificial intelligence in the banking industry is forecasted to grow significantly in the next 10 years.
According to industry forecasts, by the year 2030, the use of AI in the banking industry will generate around 99 billion U.S. dollars in value in the Asia Pacific region alone.
Fig. 1: Business value derived from artificial intelligence (AI) in banking industry worldwide
2018 to 2030, by region
Source: Statista
AI is set to have a truly positive impact to people – but it won’t come without its own set of challenges. As of late, we see many banking organisations launching their respective AI initiatives in order to be more efficient. By implementing intelligent automation, manual repetitive tasks are being reduced, interactions with customers are being improved, and sustainability initiatives are getting further along. The key challenge has always been the know-how, and to a certain degree, people’s biases around AI, mainly due to lack of understanding and wrong impressions propagated through media and other forums.
AI is set to have a truly positive impact to people – but it won’t come without its own set of challenges.
AI in 2020
Whether we like it or not, AI is here and will continue to impact people’s lives. It will bring new opportunities for businesses and the workforce alike. It will create new experiences for customers that are more aligned with their expectations. Some of the notable trends we see in 2020 are:
Open source frameworks will grow
Thanks to open source software, the barrier to entry in AI is lower than ever before. Nowadays, there are a number of open-source tools to choose from, including Keras, Microsoft Cognitive Toolkit, and Apache MXNet. Before these, Google has already warmed the market with it released its TensorFlow machine learning library as an open source back in 2015.
Predictive maintenance algorithms will gain traction
Cost pressures will continue to drive many companies to find ways to save millions of dollars in unexpected failures. As such, predictive maintenance algorithms, which constantly collect data to predict equipment failures before they occur, will gain traction. With continued drop of sensor costs and increasing availability of required skillsets, investments in this area will grow dramatically.
Auto claims processing and settlement will improve
In the past, humans manually analyse accident images and review driver behaviour based on available documentary record. This will change. With many insurers and fintech companies now using AI to compute a car owner’s “risk score”, expect the whole system to be overhauled resulting to faster claims processing and better customer experience.
Language translation will become more responsive to business needs
Natural language processing for the purpose of language translation accurate enough to respond to business needs has always been a challenge. Thanks to companies such as Google and Baidu, this might be a thing of the past. With much resources being poured into improving translation frameworks, expect great improvements in language capabilities, allowing wider adoption across industries.
Proactive cyber threat monitoring
The financial industry has been rocked with multiple cases of breach in cybersecurity. It resulted to an immeasurable amount of loss in revenue, not to mention the trust and confidence of many customers in these institutions. With advancements in computing power, proactive hunting of threats using machine learning is now possible and will continue to grow as more businesses realise its intrinsic value.
Proactive hunting of threats using machine learning is now possible and will continue to grow as more businesses realise its intrinsic value.
The China Factor
By now, China’s ambition to lead the world in the area of artificial intelligence is no longer a secret. It is determined to outpace not only the US but every global economy in AI. According to Kai-Fu Lee, a known AI expert and a Taiwanese-born American computer scientist, businessman, and author, China is now producing more than 10x more data than the US. By all indications, China looks set to catch up to the US by 2025 and lead the world by 2030 in terms of technical capabilities, available talent and most of all, data.
Most of these capabilities are coming together, thanks to China’s big tech companies who are bringing their AI capabilities to the global marketplace – Baidu, Alibaba and Tencent (collectively called BAT).
These three Chinese big tech companies combined are positioning themselves to be the AI platform of the future, giving its American counterpart, GAFA (Google, Amazon, Facebook and Apple) a run for its money.
With its breadth of resources, Baidu, Alibaba and Tencent are now able to compete with US tech giants for global AI talent. Reports reveal that “Baidu USA was offering a base salary of $130K to $175K per annum for a machine learning engineer…according to a petition filed with the US labor department for non-immigrant workers. (For comparison, Google was offering around $110K for an ML engineer, although experience levels and job requirements likely vary.)”
In addition, many of the BAT’s AI products are competing directly with what GAFA are working on such as smart speakers, AI in e-commerce, autonomous vehicles and more.
But what’s interesting to note is that Baidu, Alibaba and Tencent has strong support from the Chinese government. According to CBInsights, “China wants to be a world leader in AI in the next decade, and BAT is crucial to helping it get there.”
The Chinese science ministry announced that “the nation’s first wave of open AI platforms will rely heavily on Baidu for autonomous driving, Tencent for AI in healthcare, and Alibaba for smart cities. Government support for and intervention in AI development will likely have an immediate impact on the fast-growing Chinese tech market (where an entire city is being built from scratch around AI-centric solutions),” the report said.
While Baidu, Alibaba and Tencent are operating on a much bigger scale, the number of AI startups and corporations making headways in various sectors continue to climb. Recent figures reveal that the number of AI corporations in Beijing amounted to 422 while the number for Shenzhen totalled 167 as of May 2019.
In addition, there are a number of well-funded Chinese AI startups 2019. As of June 2019, SenseTime received total funding amounting to around 2.64 billion U.S. dollars, holding the top rank as the most funded artificial intelligence (AI) startup in China. Among the top five most funded AI startups were Megvii, UBTech Robotics, AIWAYS, and Horizon Robotics.
Across many industries, China, and by extension, Chinese AI companies, are emerging as a formidable force in artificial intelligence globally.
By all indications, China looks set to catch up to the US by 2025 and lead the world by 2030 in terms of technical capabilities, available talent and most of all, data.
How Banks Can Win Through AI
Every bank, in some shape or form, has been impacted by AI. Be it back, middle or front office, nothing has been left untouched. It’s probably common for many banking customers to have interacted with a customer service chatbot but what may not be as apparent is that many financial institutions have probably used complex machine learning to detect money launderers or sifted through mountains of data to ward off fraud and other anomalies before they wreak havoc, which is much more difficult to control.
Here are the areas where banks can win big today through AI.
Customer Experience and Front Office Support
Client’s customer support expectations are always evolving. While many Millennials will not visit a brick and mortar bank branch for anything, their expectations of its digital representative is a different story. Artificial intelligence has brought on a lot of these changes. With AI-enabled chatbots and voice assistants, information sought by customers can now be dispensed faster, cheaper and more accurately. We’re also seeing AI helping biometric authorisation as well as ‘physical’ help for those who fancy an occasional bank branch visit through an AI-enabled robotic assistant.
AML and KYC Compliance
AI can help detect fraud, identify patterns and “find the needle in the haystack”, in a manner of speaking. It can help in: transaction monitoring (identify non-obvious connections between individual transaction chains); entity resolution (creates a single unified view of a customer across local and international databases); identifying ultimate ownership (extract information to identify who has ownership or management stake); and media monitoring (categorise news articles and generate a match relevance score).
Risk Management and Lending
AI has the capability to objectively assess information and come up with equitable credit underwriting. By achieving a credible credit scoring and lending efficiently, AI can help banks assess and manage risks and how they build and interpret contracts. This is good news as there is a tremendous upside in this proposition in terms of new business opportunities and promoting people to do higher value work engagements.
Outstanding Machine Learning Initiatives
As they say, it is not a matter of if, but a matter of when and how. Many notable banking institutions are now in the thick of implementing machine learning initiatives as part of their overall artificial intelligence strategies. From stories about automation and how it would augment human capabilities, to fintech collaborations, to how banks are now using machine learning to provide frictionless, 24/7 customer interactions, there are plenty of case studies to learn from.
Taipei Fubon Bank
Taipei Fubon Bank has 4.11 million regular account holders and around 390,000 wealth management customers. Every year, some 1% of regular deposit customers get “promoted” to wealth management status through various methods and gain access to the services afforded to wealth management customers. To come up with a more efficient and scientific method to identify the potential value of customers and boost their activity with the bank (leading to higher status), Taipei Fubon Bank resorted to machine learning technology to predict customer value.
Because the value of customers has been underestimated, many customers were not cultivated by the wealth management team and only had access to the services and experiences made available to regular account holders, focused primarily on online services such as online banking and mobile banking. They did not receive personalised wealth management services or investment recommendations.
After the value of customers was identified more accurately, the customers who were promoted to wealth management status received priority treatment in offline (branches) and online channels. Offline, they were offered exclusive financial planning services supported by dedicated financial consultants, received invitations to VIP events, and receive gifts when visiting a branch. Online, they were sent regular newsletters with the latest investing, product and market information. They also received personalised product recommendations, wealth management perks and special offers. These multitude of online and offline activities and benefits forged a powerful omnichannel VIP experience for customers.
Taishin Bank
When it comes to enhancing customer engagement, relying only on traditional, structured customer data, from demographics to transaction history, has its limits. Unstructured conversation data from call centers reveal opportunistic life transitions, such as starting a family and buying a house are defining moments of the human experience, that would complete the whole personalised customer journey.
Beginning as a cross-business-unit collaboration, Taishin Bank developed and applied speech-to-text and text-mining technologies across business units – from call center to business intelligence development, digital banking, payment services, and information technology. They leveraged speech-to-text and text-mining technologies to create a hybrid data source, composed of static and dynamic, structured and unstructured data. Collectively, the bank increased work efficiency at the customer call center through the automation of call type labeling. They’ve also enhanced customer experience of Taishin’s web and mobile applications by resolving trending issues in customer calls. Furthermore, the monetisation of hybrid data is increasing due to its life storytelling marketing strategies by proactively reaching out to each customer at the point of their life transitions.
Bank of Ayudhya Public Company Limited (Krungsri Bank)
“Analytics-Based Decision Platform” is an initiative of Krungsri Bank that brings in the most efficient approach for providing the best recommendation tailored for each Krungsri customer. Leveraging on machine learning algorithm, the platform produces smart prediction and consistent recommendation for interacting with each individual customer across all touchpoints. With the efficiency of this data-driven platform, Krungsri has developed a positive customer experience that ultimately wins the customers’ long-term engagement.
As behaviours and lifestyles of each customer will change over time, they need a bank that is capable of learning and adapting to their needs. This platform was built to continuously learn from all customers and produce insights required to meet and go beyond customers’ expectations.
Image: Sean Xu / Shutterstock.com
