Artificial intelligence has arrived in personal finance with two opposing narratives, both wrong. The first promises AI will make you rich: just ask it what to buy and follow along. The second says don’t trust anything a machine says about your money. As almost always, the useful truth sits in between, and understanding it is the foundation of this whole course.

The core idea I want you to take from this first chapter is simple: AI is an extraordinary financial copilot and a terrible pilot. It knows a great deal, and it organises, explains and speeds up your work like no tool before it. But it cannot — and must not — make your decisions. Let’s look precisely at where each limit lies.

AI is a copilot, not a pilot

A copilot helps you fly better: reads instruments, warns you, prepares the route, does the heavy calculations. But the responsibility and the final decision are yours, because you’re the one who knows the destination and bears the consequences.

With your money it’s exactly the same. AI can read a fund’s factsheet in thirty seconds, organise a year of your spending, explain what APR is, or draft a savings plan. That’s copilot work, and it does it beautifully. What it can’t do is know how much risk keeps you up at night, what plans you have for five years out, or whether you’ll need that money sooner than expected. That information — and the decision built on it — lives in you.

Those who understand this distinction use AI to reach better-informed decisions. Those who ignore it hand over the controls and, sooner or later, crash.

Where AI truly shines

There’s a set of tasks where AI is today one of the best tools that exist for your finances:

  • Organising and categorising. Feed it a bank statement and it turns it into understandable spending categories in seconds. We’ll cover this in detail in block 2.
  • Explaining jargon. “What is an accumulating fund?”, “what does TER mean?”, “explain this clause to me as if I were fifteen.” AI translates financial language into something human.
  • Summarising long documents. Fund factsheets, mortgage terms, reports. It extracts the essentials and flags what to look at (always cross-checking, as we’ll see).
  • Comparing in a structured way. “Make me a table comparing these three savings accounts by interest, fees and requirements” from data you provide.
  • Acting as a sparring partner. “What questions should I ask the bank before signing?”, “what might I be missing?” AI is excellent at generating angles you hadn’t thought of.
  • Drafting and planning. A savings plan, an initial budget, an email to reclaim an unfair fee.

The common thread: in all these tasks AI processes, organises or explains information that already exists. It doesn’t predict the future or decide for you. There it’s unbeatable.

Where AI is dangerous

The same system that summarises a factsheet for you can land you in trouble if you ask it for what it shouldn’t give. These are the red zones:

  • Predicting markets. If you ask whether stocks will rise or which share to buy, it’ll give you a confident-sounding answer. That confidence is an illusion: nobody predicts markets reliably, and AI doesn’t either. Its answer sounds convincing precisely because it’s designed to sound convincing, not to be right.
  • Making up data. This is the most treacherous risk. AI can give you a return figure, a fee, a legal article or a tax fact that sounds perfect and is false. This is called “hallucination,” and with money on the line it’s extremely dangerous. Every specific figure or rule must be verified at the official source.
  • Going out of date. Many models don’t know the most recent data: current interest rates, this year’s contribution limits, a law that changed months ago. They may answer with information that was correct… two years ago.
  • Giving advice with false authority. “You should invest in X” said by an AI carries the weight of a generic opinion dressed up as expert advice. And you can’t hold it accountable.

The golden rule: the more specific and consequential a figure or rule, the greater your duty to verify it yourself.

Why it’s not your financial adviser

Even as it improves every year, there are deep reasons AI doesn’t replace an adviser:

First, it doesn’t know your full situation. A good adviser asks about your family, your debts, your fears, your taxes, your other assets. AI only knows what you tell it in a conversation, and you usually tell it a tiny fraction.

Second, it has no fiduciary duty or accountability. A regulated adviser is legally answerable for what they recommend. AI answers for nothing: if its advice ruins you, there’s no one to turn to.

Third, it can carry biases. It learns from text across the internet, where bad advice, fads and noise abound. Without judgment to filter it, it can repeat dangerous clichés with total poise.

This doesn’t mean you need an adviser for everything. It means AI occupies a different place: the tool that makes you more capable of understanding and deciding, not the voice that decides for you.

How to start on the right foot

With this foundation, you can already use AI in your finances with judgment from today. Three principles:

Use it to understand, not to decide. Swap the question “what should I do with my money?” for “help me understand my options and their pros and cons.” The first seeks an oracle; the second, a copilot. Only the second works.

Always verify the specifics. Any number, fee, deadline or rule the AI gives you, cross-check it at the original source before acting. Treat it like a brilliant but scatterbrained intern: very useful, never the last word.

Keep judgment on the human side. The final decision — how much risk, when, with what goal — is yours, because you’re the only one who knows your life and the only one who’ll live with the consequences.

In the next chapter we get practical: how to talk to AI so its answers stop being generic and become genuinely useful for your finances. Because, as you’ll see, almost everything depends on how you ask.