When ChatGPT first launched, most of us used it to write emails or craft cheeky dating app replies (don’t deny it).
But lately, I’ve noticed something new – more people around me are asking it about money – and then asking me about my thoughts on the recommendations.
“How should I invest my savings?”
“Can I retire with $1 million?”
“What stocks should I buy?”
If you’ve ever been tempted to ask ChatGPT those questions… you’re not alone.
From Reddit forums to WhatsApp group chats, everyone seems to be testing whether AI can beat the market – or at least give better advice than your friend who’s “really into ETFs” now.
But here’s the real question: can you actually trust AI with your financial future?
The idea sounds exciting – a digital financial assistant that never sleeps, doesn’t judge, and gives detailed analysis in seconds.
ChatGPT and similar generative AI tools can process massive amounts of financial data, explain investment concepts clearly, and even tailor responses based on your income, goals, or risk appetite (if you tell it).
It’s like having a private tutor for your portfolio.
Singaporeans are already experimenting with this. Some use it to understand basic investing terms. Others ask for portfolio breakdowns or ideas for rebalancing their investments. A few even say ChatGPT’s advice “feels more personalised” than what they get from robo-advisors.
Even industry players are catching on. PhillipCapital and Tiger Brokers have integrated AI chatbots – “Poems GPT” and “TigerAI” – into their platforms to provide instant insights and explain stock research in plain English.
But here’s the catch: even the companies themselves include disclaimers saying don’t fully trust this.
Why?
Because generative AI doesn’t know what’s true. It’s trained on patterns in data, not facts. That means it can produce financial advice that looks confident – but is completely wrong.
Experts like MoneyOwl CEO Chuin Ting Weber and SMU’s Aurobindo Ghosh have warned that ChatGPT’s biggest risk is not that it lies – but that it sounds right while being slightly off.
And when it comes to investing, “slightly off” can cost you thousands.
AI’s biggest strength – processing tons of data and summarising it quickly – is also its biggest weakness.
Because ChatGPT doesn’t have real-time market access, it relies on past data and whatever information was last updated in its training.
This creates a dangerous illusion: confidence without accuracy. It might tell you to diversify into sectors that were “hot” 6 months ago, or quote outdated prices. And since most users don’t fact-check every number, it’s easy to mistake eloquence for expertise.
There’s also confirmation bias – ChatGPT tends to reinforce dominant opinions from the web. So if meme stocks or tech ETFs are trending, it may overemphasise them in your portfolio suggestions.
Even when it gives sound advice – like setting aside an emergency fund or using a core–satellite allocation – that guidance comes from textbook frameworks, not your personal circumstances.
ChatGPT can’t (yet) understand the anxiety of losing 20% in a market crash or the Singaporean urge to compare every investment to CPF interest rates.
At best, it’s a useful assistant. At worst, it’s a confident intern who sounds like it knows what it’s doing – until you realise it just copied someone else’s homework.
Why ChatGPT sometimes sounds smart but gets it wrong
I’ve been following how large language models (LLMs) like ChatGPT actually work for many years now, and I realised something important – they don’t know things. They recognise patterns.
LLMs are trained on vast amounts of text from the internet. They predict the next word based on what words commonly appear together in similar contexts.
So if 100 online sources say a particular stock is “a great long-term buy”, ChatGPT will naturally assume it’s a good stock – because that’s what the pattern says.
Now, if all those sources were right, that’d be fine.
But here’s the catch – as someone who’s worked in digital marketing, I’ve seen how easy it is to manipulate that pattern. People create entire networks of websites, blog posts, Reddit threads, and social media comments pretending to be neutral “third parties”, all promoting the same product or narrative.
Search engines (and AI models trained on them) then interpret this as consensus. So when ChatGPT repeats those opinions, it’s not giving you wisdom – it’s amplifying what’s most common.
And that can be untrue.
And if you’ve noticed, when ChatGPT gives “sourced” answers on Bing or Google, those aren’t its ideas – they’re just references pulled from the same internet content that anyone can publish.
So, when you rely on ChatGPT for financial advice, you’re not really trusting “AI”… you’re trusting the reliability of the blogs and posts it was trained on.
Here’s the thing – ChatGPT isn’t your financial planner.
It doesn’t know how you feel when your stocks drop 20%, or how you secretly worry about whether you’re saving enough for your parents’ retirement.
It also doesn’t ask follow-up questions to what you asked it – potentially missing out on other crucial details you didn’t think were important.
But it can be a surprisingly useful assistant – if you use it right.
Think of ChatGPT as your research buddy, not your investment guru. It’s great for:
- Learning: You can ask it to explain concepts like “dollar-cost averaging” or “bond duration” in plain English – faster than Googling 5 different blogs.
- Exploring ideas: You can feed it your goals (“I want to retire by 55”) and ask how different strategies might get you there. It can help you brainstorm options you might not have considered.
- Stress-testing decisions: You can play devil’s advocate – “What’s the downside of investing only in tech stocks?” – and let it challenge your assumptions.
But there are clear limits:
- No real-time data – It can’t tell you what the S&P 500 did this morning or whether Nvidia’s valuation is overheated today.
- No emotional intelligence – It doesn’t care if your “aggressive growth” plan keeps you up at night.
- No accountability – If it gives bad advice, it won’t share the losses.
So, instead of asking “What stock should I buy?”, try prompting it like this:
“I’m 30, earning $5k a month, and saving for a home in 5 years. What are the general investing principles I should follow?”
That’s where ChatGPT shines – teaching principles, not predicting outcomes.
If you want a quick rule of thumb:
- Ask it for frameworks, not forecasts.
- Use it to learn, not to leap.
- Cross-check everything before acting.
Ultimately, good investing is about consistency, discipline, and self-awareness – things no AI can automate for you (yet).
ChatGPT might give you knowledge, but only you can build conviction.
I’ll be honest – I pay for ChatGPT and Gemini for work and for my own investing research. It’s fast, convenient, and sometimes it explains things even better than the “finance bros” on YouTube.
But every time I ask it for money advice, I remind myself: it’s not a financial oracle – it’s a glorified calculator with good grammar.
AI can help you make smarter decisions, but it shouldn’t make them for you. Think of it like using Google Maps: it can guide you to your destination, but you still need to drive, keep your eyes on the road, and adjust when traffic hits.
If you’re serious about building wealth, use ChatGPT to learn the “why” behind financial principles – not to chase the next “hot stock”.
Combine what it teaches with timeless advice: spend less than you earn, stay diversified, invest regularly, and keep your emotions in check.
The future of finance will absolutely involve AI – but the smartest investors will be the ones who learn to use it with discernment.
So, can you trust ChatGPT with your money?
Not yet.
But you can trust yourself to use it wisely.
And if you’re looking to invest, as always, feel free to reach out for help – I’m always happy to assist you.