A businessman stands at the top of a mountain of money as he holds a large orange flag attached to a pole.
getty
Artificial intelligence is changing finance in a big way. Firms are trying to figure out how to get a competitive edge, and keep it, as previously impossible tasks and analysis become not just possible, but in many cases, easy.
Presumably, individuals are trying to do this, too. The best example is probably market forecasting. If you know a way to prompt an AI to successfully predict stock activity, Burton Malkiel’s “random walk on Wall Street” becomes – well, something else.
Here’s how writers explain it at AI Edge Daily, where there’s a roundup of related articles on financial change in the AI era.
“For those who want to invest but don’t have the time (or knowledge) to do it manually, AI is stepping in. Robo-advisors like Betterment, Wealthfront, and even some newer blockchain-based platforms now use AI to optimize portfolios based on user goals and market conditions. Instead of paying high fees to a traditional advisor, users can access diversified portfolios, automatic rebalancing, and tax-loss harvesting — all powered by algorithms. In 2025, AI doesn’t just manage investments; it adjusts them proactively based on risk profiles, financial goals, and real-time data.”
But beyond just market predictions (which is huge) there are other innovations to apply, too: banks are using AI for fraud detection, financial planning, and high-frequency trading (maybe another flavor of market manipulation, but with its own bailiwick), not to mention compliance, and customer-facing tasks.
So, in this scenario, how do companies stand out in finance?
Notes on Competitive Finance Business with AI
Forbes CCO Randall Lane interviewed a panel on just this question. Rishi Nair of Diddo, David Cushman of HFS, and Orby CEO Bella Liu discussed how things are changing, and what strategies are likely to win the day.
Nair spoke on the value of personalizing services.
I think that … the reason that we love social media companies so much is because of the incredible personalization that each one of those companies has on ourselves,” he said. “I think that as personalization gets extrapolated into different companies, and we’re able to have our own experiences, wherever we’re interacting with technology, that’s when you can start saying that this technology knows me, this technology is able to really get where I’m coming from, and having that sort of relationship with what I’m being able to use and where I’m going allows me to trust the technology a lot more. Personalization is trust, at the end of the day.”
Liu pointed out the necessity of the human in the loop, noting the fundamental flaws of fully autonomous solutions in finance.
“You have to have the human actually touch the process,” she said. “You cannot really have AI agents just go around autonomously and do the work.”
Where We’re Headed
Later, Cushman addressed questions about the current trajectory of AI in the markets.
Companies, he noted, tend to get stuck in what he called “pilot purgatory,” unable to move toward full deployment and next steps.
“It’s about between 5% and 12% that get out of that,” he said, “so we’re talking about the firms that are actually making an impact with AI. I think of those, we’re now going to see quite a significant uptick in their uptake of AI, because there’s a small group of those firms, they tend to be firms that have already been through a significant scale (of) digital transformation, and … they trust firms that they’ve done those digital transformations with, and they’re now ready to invest in the next (way).”
Scared of Mistakes
Balancing the ideas of trust and speed, the panel talked about how to bring solutions to market, how to develop projects, and ultimately, how to listen to users.
Liu tied the concept of risk mitigation to experiences she cited with users where the users want to “slow” the program down, in order to better understand its operations.
“When we work with AI, to infuse AI in some of the workflows, you enter data into a lot of different systems, and the AI actually can do a really good job, like sometimes even about 100% accuracy,” she said. “And you can (make) the AI agent run very fast, just to automate those soul-crushing data entry jobs for you.”
However, she noted, even though the speed is more efficient, in some contexts, users want it to run slower.
“In the user research, what we learn is that the users actually prefer you to slow down the AI agents, so they can exactly see how the AI is entering the data, step by step,” she said. “And they can follow along, they feel that they can intervene when there’s a potential for mistake, and then they are fully in control as a human. So the users actually are happier that way.”
“As long as there’s an agentic flow to make the hairiness not suck as much as it does, that’s where AI can be really useful,” Nair added. “I think using the AI agent after a human, or some human in the loop has said, ‘Yes, I want you to do this,’ or ‘yes, this is exactly how I want you to do this,’ and then that is done 100% … I think it’s just the human in the loop that leads to trust.”
Watch the video for the rest, including how the panel, led by Lane, ponders how companies get “alpha” with products and services, leveraging AI to its potential. This is big stuff, and will determine how the AI rush shakes out among companies. When the panel discussed the likely position of incumbents, for example, some pointed out that although these traditional firms might not be as versatile or quick to innovate, they will have an edge in terms of the auspices of an industry. Liu pointed out how the incumbents will have the “domain experience,” and, often, a customer base.
So what are the moats? What will competitive development look like? This panel helped me to build a road map, when it comes to the impact of AI in these kinds of markets. Stay tuned for more.