Can you trust AI to manage your mutual funds?: Bruce Keith on human vs AI debate

Economic Times

18 hours ago

Bruce Keith, CEO of InvestorAI, explains how AI is transforming mutual fund investing, particularly in the U.S. He urges cautious adoption in India, stressing transparency, human oversight, and diversification. AI excels in data-driven insights and behavioral analysis but carries inherent biases and lacks accountability without regulatory frameworks.
Tired of too many ads?
Remove Ads
Q. How is AI changing the game for mutual funds, particularly at the front end for consumers? What's your view?
Bruce Keith:
Q. With that kind of gap, there's clearly a need for deeper understanding in India. What kind of AI models are mutual fund companies using, and how are they different from tools like ChatGPT?
Bruce:
Tired of too many ads?
Remove Ads
Q. Are there any real-world examples where AI-led mutual funds have actually outperformed traditional ones?
Bruce:
Q. From a retail investor's point of view, how does AI help recommend mutual funds based on goals or risk appetite?
Bruce:
Q. AI-driven advisory platforms are booming. What's fueling this surge?
Bruce:
Q. A concern here is accountability. In traditional systems, fund managers are accountable. With AI, who takes responsibility if something goes wrong? And where does SEBI stand?
Bruce:
Q. You mentioned AI reduces human bias, but does AI come with its own set of biases?
Bruce:
Q. How does AI handle market volatility, especially given the unpredictability in global and domestic events?
Bruce:
Q. Any closing thoughts for retail investors who may still be hesitant about using AI?
Bruce:
As artificial intelligence reshapes industries worldwide, the world of investing is no exception. In the U.S., over 35% of mutual funds are now powered by AI, a stark contrast to just 1% in India . So, what does this mean for Indian investors? Can AI really outperform human fund managers? And more importantly, can it be trusted?In this exclusive conversation, Bruce Keith, Co-Founder and CEO of InvestorAI , sits down with Neha Vashishth Mahajan to break down how AI is changing mutual funds, from cost and performance to risk profiling and regulation.Excerpts:From a global standpoint, especially looking at the U.S. as the frontrunner in mutual fund innovation, they've been using AI for several years now. In contrast, India is still catching up. Currently, around 35% of mutual funds in the U.S. are quant or AI-driven, while in India, that figure is closer to just 1%. That's a significant gap. In the U.S., AI is used for market sentiment analysis, stock selection, strategy reinforcement, and even writing research papers. In India, we're still mostly using AI to enhance backend operations rather than drive front-end investment research, but the direction is promising.ChatGPT is useful, but for asset management firms, differentiation is key. At InvestorAI, we've built our own foundational AI, we don't use ChatGPT, Gemini, or any off-the-shelf models. Everything is developed in-house, in India, from the servers to the code. It's similar to what firms like Renaissance or Jane Street do in the U.S. To build a real edge in this space, especially in asset management, you need to develop your own foundational AI.Absolutely. You don't reach 35% market share in the U.S. without delivering results. At InvestorAI, while we don't run mutual funds, we do manage equity baskets. Since our product went live in April 2021, we've delivered a CAGR of 45%, compared to the market's 17% over the same period. That's more than double, in live trading. In India, early forays into quant strategies weren't always successful, which is why many players are still cautious.Regulations guide this process quite tightly and require firms to assess risk appetite through specific questions. That doesn't need heavy AI, it's well automated. But AI becomes valuable when comparing declared risk appetite with actual behavior. We found in a study that about 40% of people acted in ways inconsistent with their self-reported risk levels. AI can detect that gap and help investors make better-aligned decisions.AI is a hot buzzword. Every CEO today talks about it. Many businesses integrate GenAI tools for branding more than functionality. But foundational AI platforms, like ours, offer deeper value. Younger investors, in particular, are more open to trusting AI over legacy institutions. Also, AI doesn't sleep; it works around the clock, unlike fund managers. And with the ability to reduce manufacturing costs, AI allows financial services to be delivered instantly, much like ordering food on Zomato or Swiggy.A human-in-the-loop model is essential. At InvestorAI, all recommendations undergo final human review before being released. SEBI's latest circular focused on AI in the back office, not yet on investment strategy or manufacturing. But AI's biggest value lies in reducing cost and improving access, which benefits all stakeholders. As adoption grows, I expect more structured regulatory frameworks to emerge.It absolutely does. All AI models are built on human-generated data and inherit those biases. Machines reflect whatever biases are embedded in the data and algorithms. For example, ask several GenAI models for a random number and you'll get the same answer across the board, that's bias in action. The challenge is transparency: AI systems need to be clearer about their data sources and potential biases so users can interpret results more effectively.AI can't predict unpredictable political actions or wars, but it can pick up signals. For example, just before the Israel-Iran tensions escalated, our India model shifted heavily into healthcare—a classic risk-off move. The AI sensed something was off through trading volumes and patterns, even though it didn't 'know' what was coming. With the ability to process a trillion data points daily and react instantly, AI offers unparalleled responsiveness to market shifts—something human analysts simply can't match in real time.Where Wall Street goes, Dalal Street follows. We'll see more AI integration in India. But as an industry, we need to develop transparent frameworks so retail investors can truly understand AI-based offerings. My advice: try it in small amounts. Never put your entire portfolio into AI. Diversify, have a portion in active funds, another in passive, and some in AI-driven strategies. I personally allocate about a third to AI. It's a growing space, and informed participation is the best way forward.(Disclaimer: Recommendations, suggestions, views and opinions given by the experts are their own. These do not represent the views of the Economic Times)