Why AI Governance Matters to Me

Building with GenAI changed the way I think about design.

The more I worked with LLMs, RAG, agents and AI-powered products, the more I realised that creating something that works isn't enough.

The more important question became:

Can we trust it enough to use it?

That question led me into AI governance.

AI governance is not just compliance

When I first encountered AI governance, I expected regulations, policies and checklists.

I quickly realised it is much broader.

A responsible AI product needs to consider:

Risk → Data → Transparency → Human oversight → Security → Evaluation → Monitoring → Accountability

These are not things that should be added after the product is built.

They influence the product from the beginning.

Why this matters to UX

Consider an AI assistant that recommends a supplier.

From a UX perspective, I need to think about:

• Why did the AI make this recommendation?

• What information influenced it?

• Can the user verify the result?

• What happens when the AI is uncertain?

• Can the user override the recommendation?

• Is there a human involved in a high-impact decision?

• Can the decision be audited later?

These are experience questions, but they are also governance questions.

That's where I see a strong connection between my UX background and AI governance.

My governance learning

I've been building my knowledge around established AI governance frameworks and standards, including:

ISO/IEC 42001
Understanding how organisations establish and operate an AI Management System.

NIST AI Risk Management Framework
Understanding how to identify, assess and manage AI risks across the lifecycle.

EU AI Act
Understanding the emerging regulatory requirements for organisations deploying AI in Europe.

I've also been developing my knowledge of AI risk, responsible AI, model evaluation, human oversight, transparency and accountability.

Certification

I completed the ISO/IEC 42001 Lead Implementer certification, passing with 83%.

For me, the certification isn't the destination.

It gave me a structured way to think about something I was already encountering while building AI systems:

How do we move from an AI prototype to an AI system that an organisation can responsibly operate?

Where my skills come together

This is where I see my profile becoming distinctive.

UX & Service Design
Understand people, workflows and organisational needs.

↓

GenAI Engineering
Understand how LLMs, RAG, agents and AI applications actually work.

↓

AI Governance
Understand risk, responsibility, controls and regulatory requirements.

↓

AI Product Design
Bring these together into experiences that are useful, usable and trustworthy.

Governance should be designed in

I don't believe responsible AI should be a final approval step before launch.

It should influence:

What we build → How we build it → How people interact with it → How we evaluate it → How we monitor it

This is particularly important in enterprise environments where AI can influence financial, procurement, HR, customer and operational decisions.

My biggest learning

The future of AI isn't just about making machines more capable. It's about making their use more trustworthy.

And I believe designers have an important role in that.

We sit close to the people using these systems.

We can help translate abstract principles like transparency, human oversight and accountability into experiences people can actually understand and act on.

That is why AI governance has become an important part of my journey from UX → AI Product Design → GenAI Engineering → Responsible AI.