skip to log on skip to main content
VoiceOver users please use the tab key when navigating expanded menus
Article related to:

Innovation

The Future of Financial Institutions in the Age of ChatGPT and Generative AI

Financial Institutions Group

2023-09-15 04:30

The ability of generative artificial intelligence, or Generative AI, to ingest vast amounts of data, use reason and respond appropriately to prompts lends itself to several valuable use cases in the financial services industry. From call centre analytics to fraud detection, compliance monitoring and empowering advisors with relevant insights, the technology offers almost revolutionary potential for banks – and bankers.

As its pervasiveness and potential to drive business value grows, generative AI will become a tool organisations need to embrace to maintain a competitive advantage, according to an expert panel assembled by ANZ.

Speaking at a customer event in May hosted by ANZ’s Financial Institutions Group (FIG) in Singapore, Marco Maimone, Technology Centre Leader at Microsoft, said ‘traditional’ chatbots require significant human effort to define a conversation or topic to make it work.

Generative AI-powered chatbots, on the other hand, can “Engage in a natural conversation about a broad range of topics with little to no human intervention in terms of the actual training, just the guardrails”, Maimone said. “and that’s the magic of it.”

Maimone was joined at the event by Dr. Michael Kollo, CEO and Founder of AI advisory firm, Evolved Reasoning, where they discussed the possibilities and limitations of ChatGPT and generative AI.

"Generative AI can also be combined with other models such as customer 360 to offer more extensive capabilities. For example, relationship managers can get a more holistic picture of their customers’ lifecycle and use these insights to compose personalised communications or recommend relevant products based on what’s happening in the customer’s life."

“This idea of multiple models is really important, because you’re using GPT to create the content and to ingest the information, but using other models to make decisions,” Maimone said.

Figure 1:

A High-Level Timeline of AI Innovations

{CFINFOGRAPHIC: the-future-of-financial-institutions-1.png}

Putting in place a modern AI platform and building solutions on a strong foundation of data governance and privacy can help organisations ensure the quality and reliability of their data.

Potential & Limitations

ChatGPT’s human-like intelligence has generated both excitement and alarm around the world. While the tech appears to have life-like qualities and a mind of its own, experts agree generative AI technologies work best under human guidance.

“It’s important to understand AI models don’t directly feel, they have no ‘will’, they have no opinion, and they infer the truth,” Kollo said. “Even though people get worried about how life-like they seem, they don’t have an independent desire to do things, unless you set it in a particular pathway.”

GPT can be thought of as a “document completion engine” Maimone explained, one that relies on the context humans provide through prompts to generate relevant content.

Human conversations take place in a flow and that’s what these systems are good at emulating, Kollo said.

“In fact, you get the best results from these systems when you have an ongoing conversation with it back and forth,” he said.

Optimisation

In the process of identifying use cases in financial services, it’s important to consider an AI model’s limitations aside from its potential, according to Maimone.

Commonly used Large Language Models are not designed to process numerical calculations in real time or replace sophisticated risk models. That means it is not suited for tasks requiring arithmetic, unless it is connected to another model that can do the calculation.

All organisations should ensure human supervision when using it to send automated communications, the experts said, and avoid use cases that involve regulatory oversight such as risk modelling or credit scoring.

According to Maimone, generative AI is best suited to tasks that involve “ingesting, transforming, creating content – that is the sweet spot for these use cases”.

Further, to ensure the identified AI solutions can be successfully integrated and scaled up in an enterprise, organisations should ensure several critical elements are right, he said. The first is to establish a clear vision of desired outcomes from using AI.

Figure 2:

Key Elements of Scaling AI Solutions

{CFINFOGRAPHIC: the-future-of-financial-institutions-2.png}

While generative AI brings significant opportunities, it’s critical to understand the pitfalls associated with misuse, Kollo said. In this, strong ethics serve as essential guardrails to help organisations effectively manage risk.

These guardrails include making sure content produced is authentic and accurate, and recognising that ultimately the responsibility for content generated lies with the people who created it.

According to Kollo, adopting a client-first approach is also key. That means factoring in client needs when considering possible solutions to avoid “generative spam”, and being transparent with users in instances when they are engaging with AI-generated information.

As organisations come to terms with the impact of generative AI, one thing is certain, Maimone said - “we are entering a brand-new paradigm of computing … and generative AI is not going away”.

What lies ahead

According to Maimone, AI is expected to influence all facets of our lives, functioning as co-pilots and augmenting human ability. On that note, learning how to manipulate these technologies is now a core skill to stay competitive.

A willingness to experiment is the key to getting started on the journey to integrating generative AI into everyday life and work, the experts noted.

Figure 3:

ANZ CTM Event Attendees View on Generative AI

{CFINFOGRAPHIC: the-future-of-financial-institutions-3.png}
Source: PollEv results from ANZ Conversations that Matter event, May 2023

But individuals need to be mindful of the information they’re providing ChatGPT, Maimone said, especially in a work context, to safeguard any potentially confidential information.

Kollo said a bottom-up approach can be effective, as with the adoption of other technologies like digital assets. Individuals can begin by thinking about pain points at work, finding small use cases and using the systems to solve increasingly difficult problems.

“It’s a wonderful opportunity for people to step up in their organisations and to take a lead on the subject,” Kollo said. “This is where careers and new billion-dollar companies are made.”

As Yeekei Chan, Head of Financial Institutions Group, Singapore noted in his opening address, Generative AI impacts us here now, but also our children and the future that they will take carriage of. “Similar to the long-term shifts in the geopolitical landscape, Generative AI has the potential to transform many aspects of both our personal and professional lives.”

As the tech gains adoption, it will become necessary to formulate a company-wide AI strategy to enable optimal performance and effective risk management.

Contacts

{CFINFOGRAPHIC: the-future-of-financial-institutions-4.png}

anzcomau:article-hub/topic/innovation,anzcomau:article-hub/topic/technology,anzcomau:article-hub/industry/financial-institutions,anzcomau:article-hub/campaigns/institutional/financial-institutions-group-newsletter
The Future of Financial Institutions in the Age of ChatGPT and Generative AI
ANZ experts
Financial Institutions Group
2023-09-15
/content/dam/anzcom/images/article-hub/articles/institutional/2023-08/the-future-of-financial-institutions.jpg

 

Sign up
Icon of ANZ logo coming out of an envelope

Receive insights direct to your inbox

 

Related articles

This publication is published by Australia and New Zealand Banking Group Limited ABN 11 005 357 522 (“ANZBGL”) in Australia. This publication is intended as thought-leadership material. It is not published with the intention of providing any direct or indirect recommendations relating to any financial product, asset class or trading strategy. The information in this publication is not intended to influence any person to make a decision in relation to a financial product or class of financial products. It is general in nature and does not take account of the circumstances of any individual or class of individuals. Nothing in this publication constitutes a recommendation, solicitation or offer by ANZBGL or its branches or subsidiaries (collectively “ANZ”) to you to acquire a product or service, or an offer by ANZ to provide you with other products or services. All information contained in this publication is based on information available at the time of publication. While this publication has been prepared in good faith, no representation, warranty, assurance or undertaking is or will be made, and no responsibility or liability is or will be accepted by ANZ in relation to the accuracy or completeness of this publication or the use of information contained in this publication. ANZ does not provide any financial, investment, legal or taxation advice in connection with this publication.

Top