Artificial intelligence is everywhere right now. Whether it’s ChatGPT, Claude, Gemini, or one of the dozens of other tools flooding the market, people are using AI to answer questions, solve problems, and make decisions faster than ever before. In many ways, it’s one of the most impressive technologies we’ve seen in decades.
I use AI every single day. It helps me organize information, summarize documents, brainstorm ideas, and think through complex planning situations. There is no question that it makes me more productive. In many cases, it allows me to accomplish in minutes what would have taken hours only a few years ago.
Despite using it constantly, I still don’t trust it.
That statement surprises people. If AI is so powerful, why wouldn’t I trust it? The answer is simple. AI is not an expert. It’s an intern. A very fast intern. A very helpful intern. An incredibly confident intern. But an intern nonetheless.
The problem is that AI doesn’t present itself like an intern. It presents itself like an expert. When it gives an answer, it rarely sounds uncertain. It rarely says, “I’m not sure.” Instead, it delivers information with confidence whether it’s correct or not. That confidence is what gets people into trouble.
When it comes to your finances, confidence without accuracy can be expensive. A bad restaurant recommendation might cost you twenty dollars. A bad tax decision can cost you thousands. That’s why I believe AI is an incredible tool, but only when it’s used correctly.
Over the last several years, I’ve developed five simple rules that help me get the benefits of AI without falling into the traps. These are the same rules I follow as a financial advisor and the same rules I would recommend to anyone using AI to help with financial decisions.
The Problem With Treating AI Like an Expert
I learned this lesson while working on a project that had nothing to do with money.
Over the years, I had accumulated nearly two terabytes of photos, videos, and personal files spread across hard drives, cloud storage accounts, and old computers. It was a mess. I wanted a better system, so naturally I turned to AI for help.
At first, the suggestions seemed great. It recommended folder structures, naming conventions, and organizational systems that sounded logical. Some of the ideas were genuinely useful. Others were completely wrong.
In several cases, AI recommended file paths that didn’t exist. It suggested organizational systems that would have made finding files more difficult. It even created steps that simply wouldn’t work in the real world. Had I blindly followed everything it suggested, I could have made the situation worse instead of better.
Fortunately, I stayed involved in the process. Every recommendation was tested. Every major decision was verified. Whenever something didn’t sound right, I challenged it. More often than not, AI would admit it was wrong and provide a different answer.
That experience taught me something important. AI didn’t solve the problem. It helped me solve the problem faster. That’s a very different role.
The same thing happens with financial planning. AI can help you understand concepts, identify questions, and organize information. What it cannot do is replace judgment. The moment you forget that distinction, you’re setting yourself up for trouble.