What AI can and cannot do with your money
There is a great deal of noise about AI and money. This lesson separates what these tools actually do well from what they are bad at, so you can use them without being misled.
What AI is genuinely good at
Section titled “What AI is genuinely good at”Explaining things without making you feel stupid
Section titled “Explaining things without making you feel stupid”This is the most underrated use. Financial products are deliberately opaque, and asking a human means admitting you do not know.
An AI assistant will explain what an expense ratio is, what happens to your pension when you change employer, or what a term in your insurance policy means — at whatever level of detail you ask for, repeatedly, without judgement.
For financial literacy, this is a genuine step change.
Turning messy text into structure
Section titled “Turning messy text into structure”Give a modern AI a wall of transaction text or a statement and ask it to produce a clean table of dates, descriptions and amounts, and it will usually do a respectable job. Reformatting is a real strength.
Doing arithmetic you would avoid
Section titled “Doing arithmetic you would avoid”“If I put aside 800 a month at 6% for 19 years, what do I end up with?” is a calculation most people will not do by hand. Ask an AI and you get an answer immediately, and you can vary the assumptions freely.
Two caveats: check the result against a dedicated calculator for anything important, and remember that the answer is only as good as the return assumption you fed it.
Working through a decision out loud
Section titled “Working through a decision out loud”“I have a bonus coming. I could clear my car loan at 9%, top up my emergency fund, or invest it. Walk me through how to think about this.”
You will get a structured framework covering interest rate comparison, risk tolerance and liquidity. Not advice — a way to think, which is often what you were missing.
Drafting the boring documents
Section titled “Drafting the boring documents”Letters to institutions, a summary of your accounts for your family, questions to ask a financial adviser. AI is very good at first drafts.
What AI is bad at
Section titled “What AI is bad at”Knowing your actual numbers
Section titled “Knowing your actual numbers”This is the fundamental limitation, and everything else follows from it.
A general AI assistant has never seen your accounts. When you ask “how is my portfolio doing?” it can only answer in generalities, or work from whatever you typed into the conversation — which is partial, and goes stale immediately.
This is why so much AI financial advice feels vague. It is not that the model is unintelligent. It is that it is answering blind.
Current prices and market data
Section titled “Current prices and market data”Ask what a share is trading at today and you may get a confident number that is months out of date, or invented outright. Unless the tool is explicitly connected to live market data, treat every specific price it gives you as unreliable.
Your jurisdiction’s rules
Section titled “Your jurisdiction’s rules”Tax and pension rules differ by country, change every year, and often depend on your residency status. AI models are trained on a snapshot and will happily give you a rule that was repealed, or one from the wrong country entirely.
Use AI to understand the shape of a rule. Confirm the specifics with your tax authority or an accountant.
Predicting markets
Section titled “Predicting markets”It cannot. Neither can anyone else. An AI that offers a confident forecast is producing plausible-sounding text, not insight.
Knowing when it is wrong
Section titled “Knowing when it is wrong”This is the dangerous one. These systems produce incorrect answers in exactly the same confident tone as correct ones. There is no wobble in the voice, no hedge. You have to supply the scepticism yourself.
The pattern
Section titled “The pattern”Look at the two lists and a rule emerges.
AI is strong on reasoning and language. It is weak on facts about you and facts about right now.
Which suggests the fix: give it your facts. An assistant that can see your actual holdings and current prices is answering a completely different question from one guessing in the dark.
That is the direction this course is heading. Module 4 covers asking questions about your own data, and Module 5 covers connecting your own AI assistant directly to your portfolio.
But before any of that, there is a question of what you should be willing to hand over — which is the next lesson, and the most important one in this module.
Practice step
Section titled “Practice step”Open whichever AI assistant you use and ask it to explain one financial term you have nodded along to without fully understanding. Expense ratio, XIRR, offset mortgage, surrender value — whatever yours is.
Then ask it a follow-up: “explain that again as if I am twelve.”
That second prompt is the one that does the work.
Next: Keep your financial data safe — what never to paste into a chat box.