In the world of payments, artificial intelligence (AI) is making waves. From Mastercard implementing AI tools to detect payment scams in real-time, to US Bank using AI for business travel management, it’s clear that AI is already leaving its mark. But before we dive into the specifics, let’s address the elephant in the room – the hype surrounding AI. It’s easy to dismiss it as just another passing trend, especially when we’ve seen similar hype around NFTs and Web3 not too long ago. However, there’s more to AI than just a fad.
So, what exactly is AI? Well, defining intelligence itself is no small task, but for the purpose of this discussion, let’s say AI refers to anything that resembles human thinking. For example, systems that make decisions on loan applicants based on certain criteria can be considered artificially intelligent. These systems essentially replicate what a human would do with a checklist and their own intuition about whether someone is the right fit. In other words, it’s like a flow chart – if the applicant meets a certain credit score threshold, proceed to the next question, and so on.
But there’s a new wave of AI that takes things further. This is where machine learning (ML) comes into play. ML allows these flow-chart-like systems to adjust themselves and optimize outcomes. For instance, a loan company’s ML system might analyze data on repayment rates for customers with different credit scores and discover that those with a score of 600 are just as reliable as those with a score of 700. This means there’s no need to charge them higher interest rates. By offering more lenient terms, the company can attract more customers and increase profitability. Sure, a human could potentially do this, but it would take them much longer. AI and ML take the heavy lifting off our hands and expedite the process.
Let’s shift our focus to AI in payments. The applications of ML in this field are similar. Research shows that many businesses are still waiting for payments from 2022 even in May of 2023. This is partly due to the complex and time-consuming nature of payment processes. But AI can change that. AI and ML tools can sift through piles of documents, verify their authenticity, and even cross-reference them with other sources. This significantly speeds up the onboarding process, allowing customers to open accounts and start transacting within minutes. Additionally, AI can identify shortcuts and efficiency savings, automate mundane tasks, and facilitate straight-through processing of payments. The ability to process massive datasets and compare variables in real-time is a game-changer.
Looking ahead, it’s important to realize that AI systems are not just futuristic concepts; they’ve been present in the finance industry for years. When people talk about AI today, they often refer to new innovations in the field, like large language models (LLMs). These LLMs have been trained on massive amounts of data and can generate convincing-sounding responses based on prompts. However, they do have their limitations and can produce answers that fall apart under scrutiny. It’s unclear what these systems offer that isn’t already achievable through ML.
Nevertheless, it’s crucial for the payments industry to maintain a realistic view of AI and its potential. While there are various pain points that AI can address, payment facilitation, cross-border payments, and fraud prevention should still remain top priorities. It’s possible that advancements in ML driven by these technologies will enhance existing systems and enable better data analysis. As always, the key is to stay focused on what truly moves the needle and improves outcomes for payments companies.