By Jessie Taylor
Artificial intelligence has moved from being a novelty to becoming an everyday financial companion. People are asking chatbots how to budget, whether they should invest more, how to tackle debt and what to do with a salary increase. For people who cannot afford a financial adviser, an AI chatbot can also seem like an attractive alternative: it is available at any hour, costs little or nothing and can explain complicated concepts in plain language.
But there is an important distinction between using AI to understand your money and allowing AI to make decisions about it. Recent research suggests that chatbots can get many financial fundamentals surprisingly right, while also showing weaknesses when circumstances become complicated or highly personal. The safest approach is therefore neither to dismiss AI nor to trust it blindly.
Here are things to consider before taking financial advice from a chatbot.
- AI can be surprisingly good at the financial basics
Research from MIT Sloan found that large language models generally encouraged sensible financial behaviour, including saving more during working years, investing in diversified stock funds and reducing investment risk as people got older. The researchers simulated the long-term effects of people following AI-generated financial recommendations and found that the advice could result in larger savings buffers for many people.
AI performed reasonably well in handling broad financial principles, and people who used it reported finding it useful for managing their finances. That makes AI potentially valuable for the financial basics many people struggle to understand, and a useful starting point for improving financial literacy.
- The quality of the answer depends heavily on the question
One of the most important lessons from MIT research is that the way you ask can affect what you receive.
The researchers found that more structured prompts containing detailed information about a person’s circumstances produced better advice. When users provided information about factors such as income, employment, savings and financial assumptions, AI’s recommendations improved.
For best results, use AI as a financial researcher. Ask it what information it needs before offering an assessment, ask it to identify assumptions, and ask it to explain the reasoning behind its suggestions. Even then, the answer should be treated as a starting point rather than a personalised financial plan.
READ: How women are shaping a technology-driven future
- AI struggles when your circumstances become complicated
The biggest weakness identified by the MIT research was not the basic financial advice, but the nuances. The models did not always respond appropriately when circumstances changed. For example, they struggled with situations involving unemployment and sometimes recommended spending cuts that were too severe. The researchers also found that portfolios could drift rather than being actively rebalanced as circumstances changed.
There is an important limitation: a chatbot can only work with the information it has. If you do not know which details matter, you may not know what information to provide.
- A confident answer is not necessarily a correct answer
Perhaps the most dangerous feature of AI financial advice is how convincing it can sound. A chatbot can produce a polished explanation, present calculations and confidently recommend a course of action. None of those things guarantees that the answer is correct.
AI systems can make errors, make assumptions that do not apply to you or produce information that sounds plausible but is wrong. If an AI tells you that a particular investment strategy is suitable, check the underlying facts. If it cites a tax rule, verify that rule with an authoritative source. If it recommends a particular financial product, investigate the fees, risks and terms independently.
- Your questions can influence the advice you receive
The MIT study uncovered another reason for caution: AI-generated financial advice can vary according to the person asking the question.
The researchers found differences in recommendations depending on users’ gender, financial literacy and previous experience with AI. In their simulations, these differences could compound over time into meaningful differences in projected wealth.
The finding demonstrates that there is no guarantee that two people will receive identical financial guidance even when their underlying circumstances are similar. The practical lesson is to make your prompts as clear and factual as possible.
The smartest approach is to use AI as your financial co-pilot. The evidence does not suggest that people should stay away from AI when it comes to money. In fact, there is a strong case for using it more intelligently.
It can explain unfamiliar concepts, help users compare options and turn complicated financial language into something easier to understand. The researchers also see AI as potentially complementary to human financial advice, particularly for people who cannot easily afford professional advice.
That points to a useful rule: use AI to become better informed, not to outsource responsibility for your money.
When the decision is large, irreversible or highly personal, bring in a human expert.
Sources: NPR | MIT Sloan School of Management | Fortune | Moneyweb
