How to (not) use AI for stock investing

Summary: AI has made investment research faster and more accessible, but speed is not the same as sound judgment. The smarter approach is to use AI to gather, organise and challenge information while keeping valuation, verification and the final investment decision firmly in your own hands.

Summary: AI has made investment research faster and more accessible, but speed is not the same as sound judgment. The smarter approach is to use AI to gather, organise and challenge information while keeping valuation, verification and the final investment decision firmly in your own hands. You have probably done it. Sometime in the last year, you opened a chatbot, typed some version of ‘which stocks should I buy’ and a part of you waited for the answer the way you would wait for a tip from a knowledgeable friend. I did it too. The trouble is not that the machine gets things wrong. It is that it sounds just as certain when it is wrong as when it is right. The five names that told me nothing Earlier this year I ran a small experiment. I asked Gemini for the top five stocks to buy right now. It offered a polite disclaimer, then named five anyway: ICICI Bank, TCS, Bharti Airtel, SBI and L&T. Fair enough. All fine businesses. Then I told it I was an aggressive, long-term investor. The list changed on the spot. Out went the steady blue chips, in came Bajaj Finance, Adani Ports, Varun Beverages, TVS Motor and Titan. Notice what happened there. The names changed because I described myself differently, not because any business had changed. The machine was matching my mood, not weighing companies. A real research process gives you the same verdict on a company whether you call yourself timid or aggressive, because the business is worth what it is worth. And here is the part that should worry you. Not one of t

This article was originally published on September 01, 2026.