Artificial intelligence (AI) is a hot topic right now, as a quick glance at the news headlines will tell you.
In May 2024, the European Council approved the EU AI Act, which is the world’s most comprehensive framework seeking to balance innovation with risk mitigation. Specific provisions will now be rolled out over the next three years.
The Act classifies products and services according to risk, and adjusts regulatory scrutiny accordingly. The higher the risk is considered to be, the stricter the rules will ultimately become.
If you’re a merchant operating in Europe, you’ve likely considered the relevant use cases for AI – and how you can harness its undoubted potential whilst remaining within regulatory parameters. Read on for some useful insights.
Riding the hype train
We invited author, founder, and ‘finfluencer’ Theodora Lau to join us for an episode of the Pay it Forward podcast to tackle AI, and how it can be implemented responsibly.
“It’s a really interesting hype, but it’s nothing new,” explained Lau. “It’s been around in the back-end of financial firms for years, automating processes and identifying different behaviours to inform fraud prevention.”
“What’s driving the hype is Generative (Gen) AI, such as ChatGPT, which has put this immense power into the hands of consumers. They can now see it actually doing something – generating text and graphics – and they’re curious to explore.”
“Can businesses afford to ignore it – absolutely not. But we also need to be careful. It’s one thing using the technology to recommend books and other purchases, but it’s an entirely different story when it comes to people’s money.”
Lau’s tone struck a chord with wider industry sentiment – perhaps best described as cautious optimism.
“AI in payments means opening new opportunities to provide better convenience and security through leveraging transactional data,” said Igor Skachkov, Chief Product Officer at payabl.
“These technologies leverage advanced algorithms, machine learning, and data analytics to automate tasks, improve accuracy, detect fraud, and optimise decision-making. But it’s crucial merchants fully understand the technology they’re using, or work with partners that can help them to move forward responsibly as the regulatory lens widens.”
Exploring the use cases
Gen AI can help merchants draw on dynamic data to generate real-time actionable insights, deliver more personalised experiences, and offer the right payments experiences tailored to a user’s financial behaviour and circumstances.
Gen AI-powered chatbots and virtual assistants can transform customer support, providing instant and accurate assistance. But it’s worth noting the EU AI Act has created provisions to tackle the risks posed by chatbots, which will require the producers of the underlying AI systems to become more transparent about the material used to train their models, so they can comply fully with EU copyright law.
But what are the other use cases for AI, if we move beyond the chatbots and Q&A capabilities that Lau refers to as “the low-hanging fruit.” One of the most compelling – and timely – use cases is powering the next generation of fraud prevention strategies.
Increasingly sophisticated fraudsters are exploiting the instant, and irrevocable, nature of real-time payments. The amount of fraud committed in the UK alone more than doubled to £2.3 billion in 2023, according to BDO’s FraudTrack report. Factors behind this spike include the growing popularity of online scams, phishing and system breaches, and authorised push payment (APP) transactions – where fraudsters trick people into transferring money to them.
Against this backdrop, harnessing AI to detect fraud comes with a number of clear benefits, and many of the newer solutions entering the market, such as AI-powered risk scoring models, will depend on the consistent, richer, and structured data that the ISO 20022 standard provides.
AI systems can now instantly analyse transactions as they take place, providing real-time fraud detection. This enables banks, fintechs, and other financial institutions to swiftly identify and respond to suspicious activities, which reduces the window of opportunity for fraudsters.
And whilst traditional fraud detection systems often flag legitimate transactions as fraudulent, which is a headache for your customers, AI minimises these false positives by learning from vast amounts of data and constantly refining its algorithms.
Importantly, AI systems can efficiently handle ever-increasing volumes of transactions without compromising performance. This scalability can ensure that fraud detection capabilities keep pace with your business growth.
Getting started
There’s a lot to think about when considering how best to integrate AI. There are ethical concerns and legal issues to navigate, which means caution is still very much needed.
A good starting point is making sure you strike the right balance between man and machine. After all, the ultimate aim of the EU AI Act is to make technology more human-centric, whilst in the Pay it Forward podcast Lau touches on upskilling initiatives to ensure you are augmenting humans, rather than replacing them.
In terms of a practical example, you could use AI to generate answers to a set of FAQs for your business, which should then be reviewed and edited by a human. Or Gen AI can be applied to analyse your existing customer service answers and provide suggestions to improve them.
You should also consider joining a merchant network, which often shares fraudsters’ tactics as they emerge, enabling you to keep up to date with the latest tricks. Instead of tackling fraud alone, you can take advantage of these databases and use the insights to better equip your fraud detection software. You’ll not only help yourself but also other merchants, creating an ecosystem that will collectively curtail the rise of fraudsters.
It's also advisable to work with a trusted payments partner who can help you assess both the opportunities and risks associated with AI. Our payment gateway integrates machine learning techniques, and we’re watching the space closely so we can help the merchants we work with to respond to future developments.