In the retail landscape, Walmart’s Sparky and Amazon’s Rufus are just two examples of AI-powered shopping assistants gaining prominence within e-commerce apps. Despite being an early use case for artificial intelligence in chatbots, this technology is only scratching the surface of its potential applications.
AI in Payments: Fraud, Compliance, and AML
Fraud detection has been a frequent area where AI finds application, especially in card-not-present transactions. By analyzing large datasets, these systems can identify patterns and red flags with remarkable accuracy. However, the balance between responsiveness and customer satisfaction is crucial.
Sometimes a decision is very clear-cut, but when it’s not, if you’re too permissive, you might let through fraudulent transactions. Conversely, being overly restrictive could alienate legitimate customers,” stated Don Apgar, Director of Merchant Payments at Javelin Strategy & Research.
While AI excels in parsing data, the nuances required for optimal fraud prevention and management still need careful tuning. This ensures that AI systems don’t make mistakes that could undermine customer trust or financial security.
Automating Compliance
Another area where AI can enhance compliance is in the realm of anti-money-laundering (AML) monitoring by payment processors. Vetting merchant accounts involves sifting through extensive data, a task that becomes increasingly challenging as volumes grow. AI tools can help gather and analyze this information to ensure rigorous scrutiny.
The volume of data required for AML compliance is overwhelming,” noted Apgar. Many banks and processors struggle with the sheer scale of data analysis needed to detect suspicious patterns.”
The evolving nature of regulatory requirements adds another layer of complexity. AI offers a promising solution, but its effectiveness depends on how well it can handle such vast and varied datasets.
Optimizing Transaction Routing
AI is also transforming the transaction routing process, making payment orchestration more efficient. With the proliferation of different payment methods, selecting the most cost-effective option is vital for both businesses and consumers. Traditionally, rules-based systems are employed, but adaptive learning and machine learning algorithms can enhance this process.
Machine learning relies on past transactions with similar attributes, which works well until a completely new scenario arises,” Apgar explained. AI’s ability to handle broader data sets makes it more robust for transaction routing.”
Promoting Customer Service through AI
In customer-facing scenarios, AI continues to play an increasingly significant role. It can streamline interactions by directing customers quickly to relevant information or FAQs, freeing up human staff for more complex queries.
Think of AI as a supercharged search engine,” Apgar added. It enhances efficiency for merchants, guiding customers directly to answers.”
Advancing Public-Facing AI Capabilities
The potential of AI is vast, and some financial giants have already integrated more advanced forms of AI into their operations. Platforms like Visa’s Agentic AI allow agents to perform tasks autonomously, making purchases without extensive customer interaction.
The concept is that a customer could say, ‘Find me the perfect anniversary gift for my wife,’ and let an AI agent handle all the details,” Apgar observed. However, there are concerns about AI potentially making mistakes or spending too much.”
To mitigate these risks, businesses must maintain backup systems to ensure that AI remains a supportive tool rather than a critical path in any transaction.
AI should never be the sole point of failure in a workflow,” Apgar emphasized. While it brings significant efficiencies, there is always a need for human oversight.”
