DATAx presents: How AI is making banks more competitive

Ahead of this year's AI & Big Data for Banking Summit, we spoke to Hari Eppanapally, vice-president at the BNY Mellon to explore the growing role of AI within the banking industry


The banking industry has one of the strongest financial motivations to embrace technological innovation. Aside from the profits which can be made by simply streamlining a process, it is also imperative that banks stay at the cutting edge of cybersecurity to protect itself against ever more advanced cybercriminals.

So ahead of the AI & Big Data for Banking Summit, one of the many tracks at this year's DATAx festival happening in New York this December 12–13, we spoke to Hari Eppanapally, vice-president at the BNY Mellon and speaker at the festival. DATAx asked him all about where he sees the role of AI in the banking industry going into the year 2019.

DATAx: What is the biggest difference you've noticed about the banking industry's attitude toward AI and automation during your time in the field?

Hari Eppanapally: Apart from the obvious ramp up in AI spending industry-wide, the biggest difference in the industry's attitude is in the evolution of the areas where banks are looking to deploy AI solutions. Most significantly, there is a greater push to use AI to create more competitive products which is a shift from using AI and automation to simply make existing processes efficient, or leveraging AI and automation to redesign businesses and reinventing business models. Another key area is the focus on AI-driven scaling, customization and the overall adaptability of businesses.

DATAx: What do you think is the biggest myth about AI being propagated round your industry?

HE: The following myths are still very prevalent in the industry:

Throw lots of data at AI and it gives you deep insights: Simply not true. A rigorous scientific method and high-quality data which is highly relevant to the questions to be answered are key.

AI and automation will replace workers whose work is manual, low skilled and repetitive: A naïve and overgeneralized statement. AI's most remarkable use cases and applications are in areas where the repetitive work has been taken away from humans and it has augmented peoples' ability to make more informed decisions by reducing the clutter of information and processing times.

DATAx: What has been the biggest boundary to integrating AI into your industry and what are you most excited by in AI right now?

HE: Within the banking and financial services sector, there are still challenges to the adoption of AI including a lack of distinct leadership, identifying the right problem sets to solve for and skills development within the existing workforce. Consumer-facing AI is the biggest boundary to integrating AI into interactive banking.

The most exciting areas for AI adoption in banking, however, are behavior analysis and the AI-driven prediction of consumer behavior with far-reaching utilities such as a dynamic extension of credit, advance prediction of economic distress and so on (Tencent and Alibaba already do this).

DATAx: One of the biggest concerns to everyone in the industry is data security. How has this worry changed your role and how has your response to cyber threats changed over the years?

HE: There are a few:

Anti-money laundering and fraud detection: Building our own custom fraud detection ML models, reducing false positives and improving overall efficiency.

Aggregating security data: Automating cybersecurity and cyber-compliance using ML, driven by big data (logs and events across millions of interfaces) aggregation and analytics.

Monitoring cyber threats and preventing cyberattacks: Iterative pattern recognition and threat visualization combined with feedback from human analysts.

DATAx: What new technologies and approaches to your role do you foresee for 2019?

HE:Spending on cognitive systems in 2018 rose by an estimated 54%. Some key drivers for 2019 and beyond are:

  • Reshoring and localization driven by cognitive automation.
  • Asia continues to emerge as a key center for technology innovation.
  • Robo-regulators and the consequent ramp up their use of advanced technologies in regulatory compliance.
  • Customer intelligence, opportunity acceleration using AI.
  • AI integration with Mobile Banking to enhance customer experience.
  • Acceleration of "chatbots" and "conversational interfaces".

DATAx: What is the biggest assumption financial institutions make when considering AI solutions?

HE: One of the biggest erroneous assumptions that financial institutions often make when considering AI solutions is that an effective solution is contingent on great algorithms and good high-quality data. The truth is, for an AI solution to be effective, the reliance on the human element is far greater than anticipated. Whether the problem set requires mimicking, deep learning or expert systems, along with large sets of high-quality data, the human input is critical, for both churning out effective models and to de-bias naturally occurring biases learnt from human-generated data.

Hari Eppanapally will be talking on Day One of the AI & Big Data for Banking Summit, part of DATAx New York, taking place on December 12–13. To attend and hear more great insights from other data specialists from the banking world, register here today.

Huawei announces it will reach  100bn value by 2018 small

Read next:

Huawei announces it will be worth $100bn by 2018