The payments industry has seen significant transformations in recent years, particularly with the surging popularity of mobile payments and digital wallets. As the industry continues to grow and evolves, businesses are actively seeking innovative approaches by leveraging new technology to maintain their competitive edge and deliver value to their customers. Among the array of emerging technologies, artificial intelligence (AI) stands out prominently with its capacity to quickly recognise patterns, analyse vast and dynamic datasets, and provide profound insights. Essentially, any process characterised by high volume, high frequency, and a relatively structured data format can derive considerable benefits from AI. This broadens the spectrum of opportunities for banks, fintech companies, payment service providers, and other players in the financial sector. According to Statista, the global AI market is expected to have tremendous growth over the next seven years, expanding from USD 200 billion in 2023 to USD 2 trillion in 2030. In 2023, there are already about 15,000 AI companies in the US, and this number is expected to continue to grow.
EDC has recently conducted an industry-wide survey, gathering perspectives of over a hundred senior payments professionals globally regarding the growing use cases of AI and machine learning (ML) in payments. 94% of respondents believe AI and ML are increasingly used to improve fraud detection, followed by personalised customer service (67%) and chatbot and virtual assistant (65%).
As of today, AI’s applications in the payments industry are predominately focused on three domains: risk management, customer experience enhancement, and payment optimisation. Nevertheless, as AI continues to evolve, we can anticipate the emergence of more applications.
Risk management is a critical component for any business, with its importance particularly amplified within the financial industry as the nature of financial operations requires thorough attention to risk mitigation and strict adherence to regulatory requirements. Here are several ways in which AI is used to enhance risk management and compliance.
Fraud detection and prevention
The rise in digital transactions has inevitably heightened the risk of fraud, particularly for online or card-not-present (CNP) transactions. In response to this growing threat, the integration of AI and ML has emerged as a compelling solution. Their strength lies in their capacity to quickly and accurately analyse vast amounts of data, employ predictive modelling, and identify unusual behavioural patterns in real time, thereby enabling instant flagging of suspicious transactions, ultimately leading to substantial cost savings while augmenting the overall customer service experience. Financial institutions, including banks and payment service providers, are prime candidates that can derive immense benefits from AI-powered fraud detection, given the substantial volume of transactions and sensitive customer data they manage daily.
KYC (know your customer)
It is crucial for customer onboarding in the financial industry to ensure regulatory compliance, prevent money laundering, and enhance overall security. AI revolutionises the KYC process by automating document verification, facial recognition, and biometrics, expediting identity validation and reducing manual efforts and operational costs.
AI-powered predictive analytics
It is undoubtedly a powerful tool, providing a comprehensive analysis of extensive datasets to anticipate future trends, behaviours, or outcomes. This capability assists businesses in making objective, data-driven decisions. Within the financial sector, it can play a crucial role in assessing an individual's credit risk by considering a holistic range of data points, whether historical or real-time, about the individual. This allows financial service providers to make instant yet informed decisions regarding loan underwriting.
The AI tool, while highly efficient, is not infallible, and it can occasionally generate false positives (flagging a legitimate transaction as fraudulent) or false negatives (missing actual fraudulent activities). These errors can result in financial losses, damage a business's bottom line, erode customer trust, and tarnish the overall customer experience. Finding the right balance between minimising false positives and false negatives is critical, and achieving this may involve incorporating human resources and expertise. Additionally, algorithmic decision-making can be biased and potentially lead to discriminatory practices and unjust treatment, especially for credit risk underwriting. To address this, it is crucial to conduct an ongoing and thorough audit of the AI model and promote transparency and interpretability in any AI-based decisions.
It is important to note that while businesses are leveraging AI to combat fraud, fraudsters themselves are also becoming increasingly sophisticated, utilising AI to enhance their fraudulent techniques. To stay ahead, businesses need to continuously evolve their defence mechanisms, including staying updated with the latest developments and fostering collaboration within the industry to share insights and best practices in combating fraud.
Customer experience enhancement
Chatbots and virtual assistance
In today's fast-paced 24/7 world, consumers have grown accustomed to speed and convenience in every aspect. They desire the ability to reach out to companies for inquiries at any time, rather than being confined to regular business hours. Instant replies are also part of the expectation. Even a five-minute wait/hold is considered too long. To effectively meet this demand in a cost-efficient manner, businesses are turning to AI technology, which has already been deeply integrated into customer service and support across various sectors via chatbots and voice bots. According to DataHorizzon Research, the chatbot market is projected to expand at a CAGR of 22%, reaching USD 32 billion by 2032.
Personalisation is critical in delivering a strong customer experience. AI enables companies across different sectors to delve deeper into customer behaviours and preferences, empowering them to provide personalised solutions, recommendations, and enticing promotions or incentives. For instance, in the banking sector, AI's analytical prowess allows for a comprehensive review of a customer's payment and transaction history. This analysis leads to tailored offerings, including personalised loans and credit card choices, complete with loyalty programmes that resonate with the individual customer's preferences.
While chatbots and virtual assistants have undoubtedly transformed the way businesses engage with customers, they face significant critique for lacking human empathy – an essential aspect of customer service, particularly in complex and emotionally charged situations. For instance, I recall expressing a concern to a merchant via a chatbot. It struggled to comprehend my issue – and consistently provided the same cold and impersonal responses. Frustrated, I ultimately chose to call customer service, enduring a wait of over half an hour to speak with a representative. This episode exemplifies a painfully inadequate customer experience.
Furthermore, the utilisation of sensitive customer data for personalised promotions and services raises concerns regarding data privacy, security, and the potential for discriminatory practices. Striking the right balance between automation and preserving a human touch while effectively managing data privacy and security, along with careful attention to other ethical considerations, is key to building long-term trust with customers.
Smart routing for optimal transactions
This involves predictive analysis of patterns and trends utilising the transaction amount, location, time, historical data, and payment method. AI and ML algorithms can dynamically adjust to changing conditions, effectively directing transactions to the most efficient payment networks, gateways, or processors, ultimately leading to minimised processing times, reduced costs, and an overall enhancement in efficiency.
Automating reconciliation processes
AI also plays a pivotal role in automating reconciliation processes. It has the capacity to automatically match transactions with their respective records in the financial system, swiftly identifying discrepancies and exceptions. This drastically minimises the need for manual efforts, thereby ensuring a streamlined and smooth payment experience for consumers.
The AI model has the potential to inadvertently incorporate biases present in historical data, resulting in unjust or inefficient payment routing. Moreover, when the model encounters a novel scenario, it could make erroneous routing decisions. Therefore, it's important to emphasise that human oversight cannot be entirely replaced.
The opportunities listed above are not intended to be exhaustive. Other opportunities may include the automation of currency conversion and compliance checks in cross-border payments. AI can be used to forecast currency fluctuations, allowing businesses to make informed decisions regarding foreign exchange. Additionally, it has the potential to improve the online shopping experience by suggesting complementary products or offering discounts to customers based on their browsing or purchase history.
While the opportunities with AI can be limitless, this also comes with several risks and challenges. According to our survey, the top three risks associated with AI and ML adoption are (1) lack of transparency or clarity on the decision-making process, (2) algorithmic bias, and (3) data privacy.
In summary, the impact of AI and ML on the payments industry is profound and far-reaching, encompassing improved risk management, personalised customer experiences, and payment optimisation. As AI and ML continue to progress, we can anticipate further innovations, ultimately benefiting both businesses and consumers through heightened security, efficiency, and convenience.
Embracing these technologies isn't just an option; it's a necessity for maintaining competitiveness in today's dynamic payment landscape. However, AI raises valid concerns regarding lack of transparency on decision marking, potential discrimination or bias, data privacy, and the absence of human empathy in interactions. Balancing automation while preserving a human touch and ensuring robust data privacy and security, coupled with ethical considerations, is crucial for long-term success in business.
This article was first published by The Paypers, the Netherlands-based independent source of news and insights for professionals in the global payment and e-commerce community.
The content of this article does not reflect the official opinion of Edgar, Dunn & Company. The information and views expressed in this publication belong solely to the author(s).