AI in Practice: Testing, Risk, and What I'm Actually Learning

Over the past several years, artificial intelligence has shifted from a concept discussed in boardrooms and academic papers to a tool sitting inside the apps, workflows, and devices we use every day. And with that shift has come a flood of questions — from clients, from colleagues, and frankly, from myself.

As an IT professional and business owner with years of experience in information systems, I've spent a lot of time thinking about how technology gets implemented, what can go wrong, and how to make smart decisions when the landscape moves faster than the documentation. AI is no different.

So I'm starting a series. I hold two positions simultaneously — I'm genuinely enthusiastic about what these tools can do, and genuinely skeptical about the data privacy and risk implications still being worked out in real time. What this series will be is honest, practical, and grounded in real testing.

Some posts will be technical: prompt injection attacks, how AI models handle your data, sub-processors, and how to evaluate a tool's security posture before you deploy it. Others will be written for business owners and leaders who need to understand the risk and opportunity without needing to understand the engineering.

We're all figuring this out together. I just plan to do it out loud.