Trade Credit Underwriting: An Inflection Point, Not Another Upgrade

7 min read

Table of contents

1. What ICISA Credit & Surety Week 2026 made clear
2. The tension that defines the profession
3. Underwriting is at an inflection point
4. Era1: relationship-based underwriting
5. Era2: data-driven underwriting, and its blind spot
6. Era3: the Intelligence Era
7. Fourquestions for risk leadership
8. Hallucinations:a debate framed the wrong way
9. Explainability:non-negotiable
10. WhatLangano Technologies is building
11. Theunderwriter of the Intelligence Era
12. The question for 2026

Key takeaways

1. Trade Credit underwriting is at an inflection point: astructural shift, not another software update.
2. Era 3 augments human intelligence rather thanautomating processes, breaking with thirty years of data-driven Era 2 tooling.
3. Explainability is non-negotiable for any AI deployed incredit underwriting, given Solvency II constraints and internal auditrequirements.

What ICISA Credit & Surety Week 2026 made clear

Something shifted in Trade Credit underwriting in 2026, and most actors in the sector have not yet seen it.

AI is no longer theoretical. It is no longer experimental. It is already reshaping how risk is assessed, priced and monitored. And underwriting, the core of the business, sits at the epicentre of that shift. This is the observation Olivier Placca, CEO of Langano Technologies, set out in the opening of his intervention at ICISA Credit & Surety Week 2026, before the global Trade Credit and Surety community.

This article captures the strongest positions taken in that intervention. It is written for the Heads of Underwriting, CROs, CUOs and Risk Directors who want to understand, beyond the surrounding noise on AI, what is actually happening to their profession.

The tension that defines the profession

Underwriting is one of the most intellectually demanding functions in financial services. It calls for judgement, a sense of synthesis, experience and a degree of intuition. Yet it is also, and this is the paradox, one of the most operationally constrained. The underwriter spends a substantial share of their time gathering information, reconciling fragmented data, navigating multiple systems, and repeating similar analyses to produce their write-ups.

This tension has set the operational ceiling of the profession for years:

"I have observed that tension between the high value judgement and the low value manual effort."

That tension justifies the arrival of AI in the profession. Not to automate a process, but to address the friction. AI, properly applied, does not replace judgement. It removes the friction that prevents judgement from being exercised fully. The transformation starts there.

Underwriting is at an inflection point

The central idea of the intervention is clear. Underwriting is at an inflection point. Not an incremental change, not another software update, but an inflection point. The repetition delivered on stage was less a matter of style than a deliberate warning to the audience, so that the nuance would not be lost in the usual enthusiasm of sector conferences.

Why is this moment different from the previous ones? A historical reading shows it.

Era 1: relationship-based underwriting

The first era is that of the origins of the profession. Decisions rest on local market knowledge, personal networks, paper financial statements, experience and instinct. It is the era of individual judgement, anchored in direct knowledge of the counterparties.

It has been outpaced by the scale of the modern profession, which demands decisions at a pace and on a scope no individual judgement can sustain.

Era 2: data-driven underwriting, and its blind spot

The second era began around thirty years ago. This is when credit bureaus, rating models, workflow systems and internal scoring tools appeared. Olivier Placca recalled at ICISA that he was working at Experian at the time, and that this is precisely what they were building.

This era digitised files and automated workflows. It restructured access to information in a deep way. But it left one critical point intact: the cognitive process of underwriting remained centred on the human, and manual.

The underwriter was still the one who collected, interpreted and synthesised the data before drafting the rationale. Thirty years of digitisation improved the circulation of data, without touching its human digestion. This observation opens the third era.

Era 3: the Intelligence Era

The sector is entering, in 2026, a new era that changes the nature of underwriting. Many actors still imagine that AI is merely a better scoring model, an improved version of the Era 2 tools. Today's AI is not limited to a better scoring model.

It reads full financial statements instantly, extracts signals from unstructured news, detects legal risk indicators, analyses payment behaviour patterns, compares sector dynamics and identifies early warning signals. Above all, it synthesises all of this into structured reasoning.

That last point changes the nature of the profession. An AI that produces a score belongs to Era 2; an AI that produces reasoning belongs to Era 3. This shift is captured in one formula:

"This is not an automation of process. This is augmentation of intelligence."

Era 2 automated the processes around the underwriter; Era 3 augments the intelligence of the underwriter itself. The nuance is operational, not philosophical.

Four questions for risk leadership

Before the Risk Directors and CROs gathered at ICISA, Olivier Placca put four blunt questions on the table. Are we ready? Are our organisations structured for AI-driven underwriting? Are our processes designed for intelligence produced in real time? Are we preparing our underwriters to collaborate with AI, or are we stacking tools on top of legacy frameworks?

The last one is the most direct. Stacking tools is not a transformation. This is what can be observed in many organisations that believe they are moving forward because they have multiplied the pilots, when they have changed nothing in their deeper architecture.

The warning that followed is no less direct. The pace of this shift will not align with the comfort of organisations: AI will not wait for them to be ready.

Hallucinations: a debate framed the wrong way

On the topic of AI hallucinations, which blocks adoption in many financial institutions, the position defended at ICISA is sharp. AI does not, strictly speaking, have hallucinations. What actors call hallucination is, in almost all cases, a response produced in a context that is not the right one, which becomes ill-fitted rather than wrong.

The operational consequence follows. If we manage to supply the right context, if we manage to calibrate the prompts around the AI correctly, then we obtain the right decisions and the right answers. Two disciplines flow from this: context engineering and prompt engineering. They answer the misnamed problem of hallucinations.

Explainability: non-negotiable

One further requirement underpins everything else: transparency, trust, auditability. AI, in credit underwriting, must be explainable.

In a business where every decision commits the insurer and must be defensible before the broker, the internal risk committee and the external audit, explainability is not optional, it is non-negotiable. An AI that can be deployed in production is set apart from an AI that will stay stuck at the POC stage by this requirement.

What Langano Technologies is building

The line drawn at ICISA is sharp. Langano Technologies is not building another dashboard, nor another workflow, but a layer of risk intelligence designed for underwriters.

The architectural principle breaks with that of a monolithic AI assistant. Rather than a single model trying to do everything, Langano Technologies deploys specialised agents that operate as an AI underwriting team. One agent dedicated to financial analysis, another to payment behaviour, a third to legal and litigation data, a fourth to sector exposure. These agents work in parallel, cross-check the information and identify inconsistencies.

The operational consequence is concrete. Instead of spending 80% of their time gathering information, the underwriter now devotes 80% of their time to making better decisions. This inversion is what defines, on the ground, the arrival of the Intelligence Era. The underwriter does not disappear. They arrive on the file at the moment when collection and the first synthesis are already built, and their starting point is no longer the raw data but the intelligence.

The underwriter of the Intelligence Era

The position defended at ICISA on the future of the role is free of ambiguity. AI will not replace underwriters. But the underwriters who know how to use it will redefine what underwriting means.

The role does not disappear. It transforms. The underwriter moves from data processor to risk strategist, from reactive reviewer to proactive portfolio manager. The human pillars that define a good underwriter (judgement, sense of synthesis, experience, intuition) are not called into question. They are released from the cognitive load that was preventing them, until now, from being exercised fully.

The question for 2026

AI takes its place in the sequence of historical transformations of the sector. There has been digitisation, globalisation, regulatory change. AI is the next structural shift, not a separate topic, one more wave in a long series of mutations.

But the conclusion most echoed by the ICISA audience lies elsewhere:

"The question is not whether AI will change underwriting. The question is how deliberate we choose to shape that change."

This sentence moves the debate. The choice is no longer binary between equipping or not equipping, it bears on intentionality. The word, deliberate, is worth reflection by every risk leadership team in the sector. The transformation will happen. The remaining question is whether it will be endured or chosen.

Going further

To go deeper into how Langano Technologies is building the Virtual Underwriter for actors in Trade Credit, Surety and Trade Finance, you can consult the detailed presentation of the platform or speak with the team for a scoping conversation tailored to your context.