Following his predictions for the year ahead, we spoke to Ingenico’s Simon Fairbairn to gain a greater understanding of the evolving role he believes data and AI will play in the payments sector.
PaymentExpert: Firstly, how significant do you believe AI will be when it comes to increasing efficiency and security within payments?
Simon Fairbairn: AI will be significant when it comes to increasing efficiency within payments, but only when deliberately pursued and applied to use cases where it can provide real benefit.
For example, in the security space, significant progress has already been made to support prevention of real time fraud through transaction analysis and screening.
Using greater levels of technology, AI has enabled systems to learn and adapt to the data that is being processed, and over time has become more effective at identifying and dealing with fraudulent activity.
But the value of AI doesn’t stop there. Other areas equally benefit, not simply through massively increased computational power driving the speed of processing, but through AI’s learning capacity to dynamically adjust to complex situations.
A great example is the upfront screening involved in merchant on-boarding. At the heart of the process is the assessment and pricing of risk. Using AI gives greater opportunity to place the single merchant assessment in the context of the rest of the business, and adjust accordingly to each variable assessed, individually or in aggregate.
This type of orthogonal assessment that adjusts dynamically is something that is not achievable through the application of human handling, regardless of how smart or well-trained.
Look more widely and the demand for similar AI services starts to explode. The areas of security, fraud prevention and other activities are typically fragmented in organisations; a historical by-product of how such functions have evolved. However, they all draw on a relatively common pool of data and analysis to perform their roles.
With greater access to data, the application of the latest data science techniques and the use of AI services, these functions have the potential to become even more sophisticated.
PE: You describe Artificial Intelligence has having ‘enormous potential to improve the payment ecosystem’ how do you believe banks can maximise this in 2020?
SF: Like the business process re-engineering phase of the 90s, adoption of AI is all about looking first at what you already do and re-imaging how it could be enhanced or even recreated in a more efficient, more effective and more intuitive manner.
The myriad potential on offer for what the banks already do is significant and that’s without even starting on the opportunities yet unimagined and explored. Progress has already been made in the areas of fraud analysis and prevention, anti-money laundering, and even the delivery of some customer facing services such as balance checking.
Focus first on the low hanging fruit where AI can be readily adopted and the benefits are more obvious, anything where volume data is processed and the underlying decision support systems evolve.
Given the price tag can be expensive, building momentum and delivering real value is key. But if done well, opportunities to do more and reach further will grow, allowing organisations to be more speculative and explorative in their outlook.
Further improvements will surely come in the management of risk, security and fraud. But increasingly as technology improves, we should expect more services to be delivered in the customer space, using virtual or voice bots to interact more naturally with customers to process routine activity such as making payments or more complex activity such as extending an overdraft.
PE: How important do you believe it is that the sector maximises the growth of data when it comes to boosting the user experience?
SF: Data is at the heart of how services are personalised; better data with better context enables the provision of enhanced, more personalised services.
However, given the tension around data privacy, we are likely to see a more polarised and constrained outcome as the battle between those who are happy to share in the pursuit of a better experience clash with those who are less inclined to be open.
Either way, the genie is out of the bottle and the ability to use data to provide better, richer more valuable experiences will only become the norm as marketing and delivery moves from a macro stage to a hyper-individualised one.
PE: In terms of the battle against fraud and security breaches, how impactful can data and AI be for banks and payment operators?
SF: The impact data and AI could have on the battle against fraud is massive.
The industry often talks about advances and developments being made to combat fraud but remain mostly silent on the level of innovation undertaken by the fraudster community.
However, with the scale of the values in play its should hardly be a surprise that significant investment is made to find new, cunning ways to digitally attack individuals and organisations and perpetrate fraud.
Against this backdrop, the use of data and AI has an enormous role to play both reactively and real time. Without the computational power, insight generation and safeguarding offered through such solutions, the war against fraud would remain a one sided contest.
PE: Is it imperative that the retail payment sector embraces AI in order to advance to the next level?
SF: As the biggest source of transactional payments then the application of AI to support the prevention of fraud is compelling in the Retail space. Perhaps the better question is where in the retail supply chain should this effort be focused?
Deployment at any other than the very largest retailers would be sub optimal, given the scale of cost, expertise and maintenance required. A better place sits with the acquirers, banks, payment service provides who operate the platform and provide the engine for making and processing payments.
Regardless the retail community needs a voice in this area and should play a role in both driving the providers of such services whilst safeguarding what happens at the point of sale.