How can generative AI be used in the payments industry?
Strictly speaking, the use of Artificial Intelligence (AI) in business – and the payments industry in particular – is old news.
Table of contents
Companies have been using AI to summarize and find patterns in large amounts of data, to handle basic customer interactions, and to prevent fraud for years. So – why all this buzz now, and as a payments industry executive, should you pay attention to it?
The transformative power of Chat GPT
Arguably it is the coupling of the most advanced deep neural networks, machine learning processing power and pattern recognition (all hallmarks of AI for many years), with the ability to generate outputs in recognizably and usably human format that makes the buzz around Chat GPT so unique. In other words, Chat GPT offers the processing power that only a machine can harness, and yet passes the Turing test with ease – in most cases, its outputs cannot be distinguished from those written by a human. We used to learn the language of machines to harness their power – Chat GPT is making its power available to us by learning the language of humans.
While profoundly interesting from a human perspective none of this inherently matters to you as a payment executive. The real reason you should care about AI is its power to transform, as it is already doing to a myriad of industries – from testing thousands of molecules in the pharmaceutical industry, to generating perfectly usable code based on human language prompts in a variety of high tech settings.
Three game-changing use cases for generative AI in the payments industry
As with other industries, there are many potential use cases for the payments industry. In this article, I will focus in on the three I believe to be the most transformative, precisely because they leverage the combination of the unique advances in AI embodied by Chat GPT that I described above.
Real time fraud detection
Generative AI’s ability to process immense amounts of data in real time to detect patterns and anomalies, paired with its ability to present its findings in an immediately digestible format, will transform the power of fraud detection in the payments industry.
Payments innovation is consistently accelerating, and new payment methods attract fraudsters seeking to discover and exploit vulnerabilities in new systems before they are adequately addressed. Instant payments likely present the most complex challenge among the various payment types. How can financial institutions identify and stop a fraudulent payment that takes place in mere seconds?
Most other payment types provide time to undertake checks and various mechanisms to recall funds before they’ve cleared. Instant payments not only remove this precious time, but are also irrevocable. The sender cannot stop a payment from happening once underway, and the recipient can withdraw those funds immediately.
Rules-based fraud detection systems are no longer fit for purpose in the world of instant payments. The rules are defined by fraud prevention officers and have to be manually updated as threats evolve. Between the updates, fraudsters find loopholes and without the safety net provided by a time lag, fraudsters can mount large scale attacks and reap the rewards.
While generative AI is still to be applied to fraud detection software, it highlights the lightning quick actions that could take place in those precious moments as an instant payment takes place. In addition to the speed, AI can scour historic data covering millions of transactions, current customer behavior and new trends to detect fraud and stop it by blocking payments at the messaging interface.
No army of fraud prevention officers using legacy technology can do this. And even if they do identify a new pattern, they can only work reactively, setting up a new rule after the loss has already happened. In the world of instant payments, this will be too late.
In addition, AI could predict and model emerging types of fraud and be one step ahead of fraudsters. This could be achieved by detecting anomalies related to transactions, locations and devices, combining it with information about customer behaviors, and comparing it to huge libraries of existing fraud cases – at huge scale and lightning speed.
Imagine having a fraud prevention officer with the processing power of many humans, and yet capable of intuitive learning and communication only the most sophisticated of us can tap into to make the case for a novel fraud discovery? The unique competitive advantage delivered by AI in this application is the accelerated pace of innovation due to the significantly reduced risk of loss through fraud. AI has the potential to not merely transform fraud detection – AI will transform the very way we go about innovation!
Investment decision making
Presently, humans are already harnessing the processing power of technology to extract insights from the data being processed. However, in most cases, it is the human that has to do the hard labour of providing granular testing criteria, of determining how to present the data in a digestible format, and generally interacting with the machine in a way that does not come naturally to a human. Generative AI has the potential to change all that.
In this use case, decision makers would simply request the information they need, and have it presented in a form they can easily digest, along with supporting evidence to help make the best decision. In the context of payments, the generative models could also be used to react to live data and continually update dashboards for new, or fast-moving trends, enabling even faster decision making.
This could be used in several areas, but the one I would call out here would be for a generative AI model to be used to determine how a range of payment methods are likely to be used in a particular geographic region, or by a specific customer group. The model would indicate where the opportunity lies and where further investment in product, business development and marketing may be required.
For example – not all payment methods are ubiquitous in every region and market segment. In North America and Western Europe, card-based payment products (including both mobile and contactless), are still prevalent. On the other hand, only 5 to 7 percent of all payment transactions in Africa were made via electronic or digital channels, compared with 50 percent or more in Turkey, for instance. Yet, it is in the emerging markets such as Africa that true innovation and growth opportunity lies, with their ever increasing number of mobile connections, enabling leapfrogging of technologies in the near future.
Enter generative AI – with the thousands of market testing scenarios it could run without the need for advanced, granular statistical programming. Market testing an innovation that would take years may well be reduced to days – not to mention the truly transformative reduction in risk and cost!
Advanced payment product technical support
By 2025, it’s estimated that 95% of interactions between brands and customers will be powered by artificial intelligence. This is not to say that there won't still be a role for a human in such interactions. Humans will be enabled by AI to offer the right solutions to the customers, faster.
One major bottleneck to innovation adoption is the ability to train up the customer service technical support workforce to enable them to effectively support and onboard a large number of new customers, quickly. In other words, there simply isn’t enough time in the day.
Generative AI has the capacity to change all that. For example, Gen AI could be integrated into a company's customer support system, where it could help to quickly and accurately answer customer inquiries, freeing up human agents to handle more complex issues. This would in turn increase customer satisfaction, and significantly reduce friction associated with new payment product adoption. Imagine being able to provide technical support to tens of thousands of small merchants, all at once, at a fraction of the cost in time and money!
The opportunity cost of passing up on generative AI is too significant to ignore
Back to my initial question, then: should you care about generative AI as a payments executive?
The answer is that there is no certainty in this world, and yet it is our job as executives to assess risks and make informed decisions with the objective of improving the businesses we run. And with so many potentially game changing use cases, the decision we must take with regards to AI seems to be clear…
We essentially arrive at the modern version of the famed Pascal’s wager. What’s the worst that can happen if we bet on generative AI, and it turns out not to be as transformative as we thought? Our business model would be no worse off than where we started…and yet, should a competitor capitalize on the transformative power of generative AI, the consequences can be disastrous for the businesses we run.
If ever there was a decision that did not require the use of generative AI to make, it is this one – the opportunity cost of passing up on AI is too significant to ignore!
At Vega IT we have extensive experience in FinTech and payments, and our data science and AI team would be delighted to help you with your strategic considerations of using generative AI in your payment system.
Latest blog posts
Next stop in the Vega IT global expansion: The UK offic...
During the last year, we witnessed significant changes in the tech landscape. Many global companies focused more on optimizing the costs, often through layoffs, rather than investing in new projects or entering new markets....
Why AI in EdTech is the disruptor we need
EdTech is revolutionary in its mission to transform the way we teach and learn, but why does it generally "lag behind" when it comes to technological advancements? Moreover, why does it shy away from AI?