One of the biggest challenges in AI today is not finding use cases.
It is making sense of them.
At the Global AI Governance and Innovation Showcase, Huy Nguyen Trieu shared the thinking behind structuring AI adoption in financial services through the AI in Finance Observatory.
The Problem With “Using AI”
Many organisations now say they are using AI.
But that statement can mean very different things.
It could refer to:
- an early idea
- a prototype
- a pilot
- an internal experiment
- a production system
- a scaled business solution
Without clarity, it becomes difficult to understand what is really happening across the industry.
Are firms experimenting?
Are they deploying?
Are they scaling?
Or are they simply exploring?
Why Structure Matters
AI adoption in financial services is becoming increasingly complex.
Use cases can appear across many areas, including:
- customer service
- compliance
- risk management
- operations
- fraud detection
- product development
- investment processes
They can also create different types of value.
Some use cases improve productivity.
Others reduce costs.
Some create new business models.
Others improve decision-making or customer outcomes.
Without a structured way to compare them, it is difficult to identify patterns or understand what works.
From Complexity to Clarity
The approach presented was to create a framework that maps AI use cases across two key dimensions:
- where AI is being applied
- what value AI is creating
The goal is not to add complexity.
It is to make complexity understandable.
For institutions, this helps identify opportunities.
For regulators, it supports better visibility.
For innovators, it creates a clearer way to explain impact.
The Takeaway
As AI adoption accelerates, progress will no longer be measured by experimentation alone.
The next phase will require:
- clarity
- comparability
- structure
- scale
Huy’s message was clear:
To move forward, the industry needs a common way to understand what AI is doing, where it is creating value, and how it can be scaled responsibly.
