AmEx is cautiously experimenting with generative AI for fintech

Executives at AmEx Digital Labs, the company’s innovation arm, say they likely won’t launch their own LLM and instead rely on integrating existing offerings

The wave of new business products and services launched with generative artificial intelligence (Generative AI) capabilities isn’t slowing down, but many established companies are taking a measured approach to evaluating how to use and implement the technology.

Among those exploring artificial intelligence for business is American Express (AmEx), the giant of financial services, credit cards and remote concierge services. Founded in United States in 1850 originally as a physical goods delivery company, AmEx has transformed into a global fintech leader, constantly embracing new technologies throughout its nearly two centuries of existence.

Now, with the recent launch and pace of hype around OpenAI’s ChatGPT, Google’s Bard, Anthropic’s Claude, and similar generative AI products built on large language models, AmEx sees an opportunity to use these technologies to enhance the customer experience through its credit card and bank offers for businesses and individuals.

“It may not be a standalone product this year, but we’ll be looking at direct-to-consumer or small business services as part of our experimentation and see what works and what doesn’t,” said Luke Gebb, executive vice president, American Express Digital Labs . AmEx Digital Labs is a division of AmEx formed in 2017 with a startup ethos, rapidly evaluating and testing new technologies and using them to incubate new products that find their way into the rest of the company.

AmEx Digital Labs was working with artificial intelligence before it became trendy

In fact, AmEx Digital Labs has already been at the forefront of using AI for financial services. This team of 100 tech experts from around the globe was responsible for integrating Mezi, an AI-powered digital assistant that provides travel booking recommendations to customers, into a AmEx mobile app pilot program. AmEx later acquired the agency behind Means and has woven its technology into some of its services.

“We launch 20 to 25 pilots a year,” said Laura Grant, vice president of product development for emerging platforms and artificial intelligence at AmEx Digital Labs. “We look at technology from the perspective of ‘How is this going to impact our customers and improve their lives?'”

The way AmEx Digital Labs works is that they first build and prototype products within their own small unit, before distributing them to other parts of the company, finally handing over control to whichever team within AmEx is best equipped to drive that particular digital offering.

For example, AmEx Digital Labs built the company Awards Bank accounta high-yield consumer banking offering (1% annual percentage return) in which customers earn reward points for every dollar spent on their debit cards, then turn those points into direct deposits.

Additionally, AmEx Labs defended the company”Pay with Bank Transfer”, which allows online merchants to include an AmEx-powered button among the payment options they accept. If a customer hits the “Bank Transfer” button, they are asked to securely and privately link their checking account at one of the many supported banks and pay directly for the product from it, avoiding credit card transaction fees.

How AmEx plans to use Generative AI

AmEx Digital Labs intends to explore the potential of generative AI in enhancing predictive analytics, helping them better understand customer patterns and behaviors.

In relation to AI applications, AmEx aims to focus on “predicting how our customers will behave over time, enabling better financial planning and decision-making,” Gebb said. Another important use case: analyzing customer sentiment and customer interactions.

“The way I think about AI is how we think about customer experience,” Grant said. “What will it do to make people’s lives easier and improve the customer experience?”

AmEx Digital Labs is currently exploring ways in which the LLMs could be used ‘behind the scenes’ to analyze all feedback and inquiries customers provide through AmEx’s existing customer service portals, as well as unofficially on social media, and to understand how to provide appropriate information and helpful answers.

Grant said that when approaching all new technologies, but artificial intelligence in particular, AmEx Digital Labs tries to figure out how it can help with its “3 Ps,” making a product more personalized to an individual customer, more proactive and more predictive.

As for whether AmEx will look to develop its own in-house LLM using its financial services data, following in the footsteps of financial information and media giant Bloomberg which launched its BloombergGPT with 50 million parameters a month ago, Gebb was dubious.

“Our guess at the moment is that we would be more suited to using LLMs through partnerships,” Gebb said, but added that he didn’t have any specific ones to announce. “I don’t see us creating our LLM from scratch.”

Security fence

While AmEx Digital Labs sees customer-facing generative AI based on an LLM as potentially useful, for now it is focusing more on back-end implementations.

“Initially it seems smarter for a human to be looking at and reviewing the output” of an LLM, Grant said.

Additionally, due to the critical importance of its product offerings – affecting customers’ bank accounts and lines of credit – and the significant amount of associated government regulation in financial services, AmEx Digital Labs also makes sure to address all of its experiments with AI very carefully and safely.

When asked how AmEx could avoid the same fate as the three SAMSUNG employees who have reportedly shared confidential and proprietary information with LLMGebb said, “We’re fence-in,” referring to the broader one safety approach which limits the software applications and the data they can access.

“Not all employees can access the LLM via their laptop, only those who have been authorized and trained in what they are doing.”

How AmEx’s AI approach compares to that of other large enterprises

AmEx’s experiments with generative AI are clearly in their early stages, but this falls in line with many of its peers.

A recent KPMG survey found that 65% of 225 executives surveyed believe Generative AI will have a large or extremely large impact on their organization in three to five years, but 60% say they still have a year or two to go. are implementing their first solutions and plan to spend the next 6-12 months increasing their understanding of how Generative AI works, evaluating internal capabilities, and investing in new tools.

The survey also showed that the executive priority of generative AI varies significantly by industry. Most executives in technology, media, telecommunications, healthcare and life sciences felt they had adequately prioritized generative AI, while only 30% in consumer and retail said they had a priority. Respondents across technology, media, telecommunications, and financial services said research into generative AI applications is a high or extremely high priority over the next three to six months.