AI has been catching everyone’s attention over recent years to the point where to many the talk about its potential has become somewhat of a broken record.
Various firms from traditional banks to emerging challengers to innovative fintechs have been looking for ways to implement AI, but have its use cases really expanded beyond chat bots?
To gauge how the industry expects AI development to play out over the next 12 months, Payment Expert spoke to Swiipr’s Ian Clowes, Toqio’s Eduardo Martinez and Blink Payment’s Ari Eder in a roundtable discussion.
Payment Expert: Is the hype around Artificial Intelligence justified or is there still more fine tuning to do before it makes a substantial impact on payments?
Ian Clowes, Co-Founder and Chairman of Swiipr: As it stands in the payments environment, AI is still mostly being used to realise operational efficiencies rather than for mission-critical systems.
Because it’s still somewhat unreliable, it’s being utilised in areas like customer interaction via chat bots rather than payment processing systems where an error could mean consumers are left stranded at the till because their payment has been declined.
With my entrepreneur hat on, I’d say the speed of development in AI is going to be much greater in future than it has been in the past. But the ex-banker in me knows this optimism has to be set against a backdrop of increasing compliance and regulation, which could constrain the use cases for AI in some respects.
Eduardo Martinez, CEO and Co-founder at Toqio: We see AI as having a lot of potential in the payments and fintech space, but believe there is a need to move forward with caution, particularly given the tightening regulatory landscape.
AI could help transform outdated, manual processes into streamlined, efficient systems. In addition, it could help businesses unlock new opportunities, such as the ability to offer digital gift cards for payouts and rewards.
We think it’s possible that AI will drive the growth of technology evolution and customer expectations around experiences. Industry leaders could build end-to-end closed ecosystems, leveraging financial services to drive retention, growth and acquisition.
Ari Eder, Head of Product at Blink Payment: AI has proven to be an extremely powerful tool, showing significant improvement since its launch. To date, its primary applications have involved analysing large volumes of data and aiding in complex decision-making processes. On the payments front, AI has most notably been utilised in use cases, such as fraud detection and reporting.
AI has begun to evolve further more recently, becoming more generative and realistic. It is anticipated that in 2025, AI will become more mainstream within the payments industry. However, concerns about security and regulation persist. It is crucial to address how we can effectively manage AI to prevent it from malfunctioning or being manipulated, thereby ensuring the highest levels of security in payment systems.
PE: How can technologies like AI revolutionise the speed of transactions for the foreseeable future?
Ian Clowes: I don’t view the speed of transactions as being the area where AI has the most upside. Transactions are already taking place in milliseconds, and a slight improvement in this speed is unlikely to prove a game-changer. The real opportunity I see with AI is if we can harness it to add value to a transaction taking place in milliseconds.
Many virtual transactions use third-party authorisation to prevent fraud; imagine if while that was taking place, AI was mining all the data on the consumer to offer them suggestions on other things they might like to buy based on their purchasing history?
It could be that someone is buying something in Dubai Airport and while their payment is being verified, AI suggests an item they’ve been looking at for some time is on sale at the shop next door. Rather than the speed of transactions, the powerful thing AI could add is speed in processing information and data to add value.
Eduardo Martinez: With the advent of new rules requiring UK banks to reimburse victims of push payment fraud coming into play this year, we saw some innovative uses of AI emerge as providers prepared for the new regulations by introducing new fraud tools, some of which were AI-based.
While these systems may have initially introduced delays as providers refined their capabilities to meet compliance and mitigate risks, their evolution promises to streamline transaction speeds significantly by reducing manual interventions and false positives.
This shift is already evident in leading AI-driven fraud detection models currently used in e-commerce and digital banking, so we expect that enhanced fraud detection systems that incorporate AI will certainly become more widespread in the B2B space.
There’s also the idea that predictive analytics can be used to optimise networks. By providing AI with enough historical data, it can do an excellent job of forecasting peak transaction times to shift network performance as required. This reduces congestion, ensuring smoother and faster processing during high-demand periods.
Lastly, there’s a lot of great work being done with natural language processing (NLP) for query handling. Voice assistants and chatbots have fallen a bit out of favour over the last couple of years due to more advanced AI offerings. It seems like people are willing to re-engage with them for support and transaction requests if they can offer the capabilities of AI.
If chatbots and voice assistants can leverage AI to interpret and process verbal user instructions instantly, those users will enjoy quick and simple access to fund transfers, account updates, and support channels.
Ari Eder: Payments can be much more complicated that they appear at first glance, and if AI could be embedded into that process, it does have the potential to streamline and speed up transaction reconciliation.
Something that would be interesting is if AI could also be utilised to make personalised suggestions to consumers as their payments are being processed. Taking this a step further, it’s possible that AI could even be used to make purchases for consumers. In the latter case, it would inevitably raise questions about fraud and how data could be kept secure though, so this idea is one to watch for now.
