Ai in Credit Management, where do we stand ?

My answer is everywhere and nowhere !

Yes, that might sound like a classic non-committal response… but let’s be honest: everyone’s talking about AI (especially in financial services), but how many real models are actually available and for what purpose ?

By definition, Artificial Intelligence refers to “the capacity of a machine or computer system to simulate human cognitive functions (such as learning, reasoning, perception, language understanding, and decision-making) to solve problems or perform specific tasks.”
Let’s be clear: in Credit Management, AI in its strictest sense is still almost non-existent—so, we could say nowhere… for now.

Moreover, there’s a widespread confusion between automation, RPA, digitization, algorithms, APIs, decision trees, and process digitalization. This confusion is convenient, because it keeps things vague and gives the impression that AI is already in place. In reality, all of these technologies are already in use across credit management so from that angle, we could say AI is everywhere.

Let's briefly review some key initiatives : 

Credit Insurers

Leading the charge, credit insurers were early adopters of autonomous decision-making models: buyer scoring algorithms, machine learning for predicting default risk, fraud detection tools, and recovery probability models are some examples.
Most of these algorithms are “supervised,” meaning the machine makes autonomous decisions up to a certain threshold, beyond which a human gives the final approval. This enables insurers to instantly make thousands of credit decisions, leaving only high-value cases to human underwriters.

Fintechs

Innovation in trade finance is clearly driven by fintechs, rather than banks & traditional factoring players. Whatever the financing method (BNPL, factoring, reverse factoring), delivering an instant credit decision on a B2B Buyer is now possible. Thanks to APIs that interface with credit insurers (a key collateral for financing), fintechs can automatically fund transactions—regardless of the purchase channel: marketplace or physical store.

Rating and Scoring Agencies

Credit information providers are turning to Large Language Models (LLMs). Once the data is collected and structured, these models are trained to apply contextual rules and logical frameworks to deliver recommendations—such as credit scores, 12-month default probabilities, and suggested credit limits. It’s important to note that these limits remain credit opinions, and not credit limits.

Credit Management Platform Providers

In theory, these are the most data-rich players, having direct access to the holy grail: “DATA”. Connected to ERPs, receivables ledgers, business reports agencies, insurers, and even collection companies, they enjoy a 360-degree view of receivables. Their AI-driven efforts focus on building internal scoring models that blend internal data (order history, payment behavior) with external inputs (ratings, credit limit, etc.) to generate scores that support credit management decisions.

And what about brokers ? 

Two major roles stand out:

1. Advisory role : Brokers sit at the crossroads of all these stakeholders. They must be at the forefront of AI—not only to advise clients on the best solutions, but also to challenge and work around the current limitations of these models. Decisions made by algorithms must always be open to review by humans—especially brokers and their tools.

2. Tools (let’s talk about AU Group’s tool in particular). These are critical to the broker’s continued relevance in the years to come. 

   We can distinguish:

   • Credit Management platforms: comprehensive tools for credit managers to manage their entire “order to cash” process.
   • Broking tools: internal solutions for brokers to track performance of credit insurance programs, insurer’s risk appetite and early detection of room for improvements.

Their specific value lies in the real (not artificial) intelligence they offer: benchmarking insurer positions, anticipating risk withdrawals, challenging underwriting decisions, and mitigating client risk by proposing alternative coverage strategies.