Should you trust AI chatbots for investment decisions?

An AI chatbot is a computer program that uses artificial intelligence to understand what you type or say and tries to reply like a human. Popular examples include ChatGPT, Google Gemini, Microsoft Copilot, Claude, etc.

Chatbots are fast becoming part of everyday life: from finding the best hotel, planning a three-day trip, to choosing the best TV, to self-help guides, AI has taken over the imagination of many households.

Money management naturally can’t be too far behind. To understand the popularity, we asked ChatGPT how common investment-related queries from India are. The answer was predictable: given India’s high retail participation, investment-related questions rank among the most frequent and consistently high-engagement prompts.

While AI chatbots are a great tool to learn concepts, narrow down choices or even discover options, the risk lies in treating recommendations as advice. Action-oriented money decision-making driven by AI can be risky because chatbots have no accountability or fiduciary responsibility.

So when you ask your AI tool for the best mutual fund in the market, understand that the answer may be more a function of the fund's visibility, repetition and other factors rather than your age, risk profile or even your investment goal.

Read on to understand why your AI chatbot can be a great teacher, but not your financial adviser. Here, we are referring to publicly available AI tools and not custom-designed tools being used by financial professionals.

Rise in AI-led queries

Advisers say investors have already started exploring different investment queries like best mutual funds, best insurance etc. on AI chatbots. “Clients are coming with AI Chatbot screenshots to us and discussing the suggestions given by these platforms, increasingly," says Akshat Garg, head - research & product at Choice Wealth.

Aditya Agarwal, co-founder, wealthy.in, says that sometimes clients have referred to the conversation that they had particularly with the popular ones – ChatGPT or Gemini.

“Most commonly, investors bring screenshots of ‘best’ mutual funds and ‘best’ insurance according to these platforms and compare them with our recommendations," says Pramod Sharma, partner, Citrine Financial Services LLP. He adds that we had a few cases where investors performed ‘how much SIP is required for ₹X crores’ calculations on them, which were incorrect.

No doubt, AI chatbots are increasingly being used as decision companions, and companies and brands know it.

The new race: ranking inside chatbot answers

Most firms won’t openly admit they are trying to ‘rank’ inside ChatGPT answers. But the incentive is now too obvious to ignore. Most of this optimisation isn’t illegal or secretive; it is simply marketing adapting to a new discovery channel.

Semrush, the international digital marketing platform, predicts that AI-driven channels could rival traditional search in economic impact by 2027, and notes that AI-powered visits convert 4.4 times higher than organic search.

The shift is not theoretical in India. As reported in October last year, an NPCI-Razorpay-OpenAI pilot is enabling UPI payments within ChatGPT, pointing to a future where discovery and transactions happen inside chat interfaces. India is already at the centre of this change, accounting for 13.5% of ChatGPT’s global user base, according to Mary Meeker’s Trends – Artificial Intelligence Report.

Big Tech is clearly betting on India: Google has offered its $400 Gemini AI Pro free for 18 months to Reliance Jio’s 500 million users, while OpenAI made ChatGPT Go free for a year in India, a plan earlier priced at $54. These India-only moves make one thing clear: brands will inevitably optimize to be the name that appears when users ask “best X".

How brands optimise for AI

“Earlier users searched on Google. Now they’ve started asking chatbots directly," says Goutam Singh, a Noida-based AI SEO and digital marketing consultant. That shift is pushing brands to optimize not just for clicks, but for mentions inside AI answers. This is where AEO/GEO comes in: Answer Engine Optimisation (AEO) and Generative Engine Optimisation (GEO) focus on shaping content so a chatbot is more likely to recommend a brand name when a user asks “best X".

“To get your name picked up by AI answers, you first have to build trust using Google’s E-E-A-T Framework… expertise, experience, authoritativeness and trustworthiness," says Singh. He adds that social presence should be visible, author profiles should be real, and names should repeatedly appear across reputed platforms.

Brands do this by moving away from long, messy blogs to clean site structure, fast performance, clear schema, and well-organised FAQs and comparisons – formats AI systems can easily read and trust. “Repeated mentions act like a signal. So if a name keeps appearing for a keyword, the system treats it as important," says Singh.

The real danger

But the danger isn’t that chatbots are deliberately misleading. It’s that they are not designed to operate like fiduciaries.

For example, we asked base versions of ChatGPT and Gemini for the best mutual funds to invest in. While both platforms caveated the recommendations advising customers to seek professional advice regarding risk, goals, etc, the recommendations were similar.

“The problem with AI is bulk data-based recommendations. AI will pick up products that rank high on the internet without any context to an investor's need," explains Ravi Handa, founder and creator of Handa Uncle, that’s building an AI-powered personal finance platform.

“While I think AI tools are far superior to commission driven intermediaries, they are no substitute to professional advice that contextualises advice based on a person’s needs. AI tools like ChatGPT are already working on advisory models like ChatGPT for health, hopefully the same will happen for finance," he adds.

The Reserve Bank of India has already cautioned consumers in its report titled “Framework for Responsible and Ethical Enablement of Artificial Intelligence (FREE-AI)- August 2025": “When AI is used to enhance engagement, it can subtly influence consumer decisions in ways that may not always align with their best interests."

Why ‘money’ is the riskiest category

Finance is a category where a confident shortlist is often the most dangerous format. “Investors must understand that if ranking is influenced by data availability, popularity, or marketing partnerships, unsuitable products can surface as “best", even when they don’t fit them," cautions Garg.

That is why, as per Agarwal, it is always highly recommended to check the sources of the answer and verify the output before acting on it."

But for many investors, that verification step never happens because chatbots sound authoritative. “The danger is false confidence– AI sounds convincing even when context is missing or outdated," says Garg. And he adds that first-time and DIY investors are more vulnerable to act on AI’s advice without any research.

The issue becomes bigger because the questions being asked are not about definitions. They are about action. These are precisely the moments when emotion is high and errors are most costly.

The accountability gap

This is where another gap becomes impossible to ignore: accountability.

Unlike Securities Exchange Board of India (Sebi)-registered advisers or regulated intermediaries, AI chatbots sit outside the traditional fence. Sebi has taken note of AI’s role in financial markets. The regulator’s June 2025 consultation paper on responsible usage of AI/ML in Indian securities markets focuses on AI inside regulated entities – governance, controls, monitoring, and accountability mechanisms. Sebi has also discussed the need to assign responsibility when AI tools are used by intermediaries.

But who bears the responsibility if an AI chatbot suggests some funds or insurance, etc, that does not work in favour of the investor?

“Today, the investor bears the risk," says Garg. “AI Chatbots are not fiduciaries, not registered with Sebi and not accountable. That’s a major regulatory gap," he adds. The RBI’s FREE-AI Report too warns of lack of clarity around accountability if an investor is harmed.

AI chatbots: Not substitutes for registered advisers

Experts believe an adviser’s technical input and behavioural handholding is still hard to replace.

“The quality and timely handholding that an adviser/distributor provides to an investor is not replaceable," says Rushabh Desai, Founder at Rupee With Rushabh Investment Services, a Mumbai-based mutual funds distribution firm.

According to Agarwal, “Genuine advice will always require many traits of the investors to be aware of which he himself is not aware of."

While Garg believes that final decisions must still go through a qualified adviser, he says those who still want to explore, should treat AI like Google– informative, not decisive.

According to Pramod Sharma, investors can use AI to learn. “It is okay to ask questions like what is mutual funds, what is SIP, old tax regime vs new tax regime etc… They can even ask AIs to simplify newspaper articles or concepts to understand them with ease."

But, investors should strictly not ask AI chatbots to create an investment plan or portfolio allocation for them. “It doesn’t know your risk profile, goals, tax status, or liquidity needs. Hence, they cannot develop a reliable plan for you," says Garg.

Sharma says that answers from AI also depend heavily on the quality of prompts and the version you use. “Suppose you want it to calculate how much you should invest monthly towards a goal, unless you write a clear prompt hinting it to use a particular formula, or may be using a paid version, the chances of getting wrong answers is high," he says.

If you are keen, says Garg, ask questions to AI; there is no harm unless you react to it. Andhe strongly urges investors to always verify recommendations, information, and data through a registered adviser or official scheme documents.