For many investors, a company’s economic moat is one of the most important things to understand. A strong moat can protect market share, support pricing power and help a business earn attractive returns for years, even as competitors try to take those profits away.
The best moats usually come from something genuinely hard to replicate: a unique product, valuable intellectual property, scarce assets, scale or some other lasting competitive advantage.
But not every moat is built that way.
Some businesses may be protected, at least in part, by the fact that their customers are human.
People are busy. We forget things. We don’t always read the fine print. We put off switching banks or insurance companies because it sounds like a hassle. We keep paying for subscriptions we barely use. And when shopping for something becomes too complicated, we often settle for good enough.
I think of these as behavioral moats. And the rise of personal AI agents could put some of them at risk.
We got an early look at that concern in September, when Meta’s Muse AI agent caught investors’ attention. Shares of companies that could be vulnerable to declining consumer inertia sold off. Charles Schwab fell more than 6%, insurer Allstate dropped 5.5%, Planet Fitness fell as much as 11%, and travel companies including Expedia and Booking Holdings also came under pressure. (Bloomberg)
The reaction may prove overdone. Agentic AI is still young, and there are major questions around security, trust and how much authority consumers will actually hand over to software.
But the underlying question is worth considering now:
Would a company’s advantage still exist if every customer had a tireless personal assistant constantly looking for a better deal?
When switching is the moat
Financial services may be one of the clearest examples.
Consumers routinely leave money in accounts paying less interest than they could earn elsewhere. That isn’t necessarily because they believe their bank offers the best deal. Moving accounts can mean comparing alternatives, filling out forms, changing direct deposits and reconnecting automatic payments.
Most people have other things to do.
An AI agent does not. In theory, it could regularly compare savings rates, credit card rewards or brokerage products and tell its owner when there is a better alternative. Eventually, agents could handle more of the switching process themselves.
That possibility has already gotten the attention of financial companies. Schwab CEO Rick Wurster recently said the company is blocking unauthorized outside AI agents attempting to access customer accounts or move money. Schwab is instead building controlled ways for outside AI systems to connect to its platform, while limiting what they can do. Other major banks are taking similar precautions. (Bloomberg)
There are legitimate security reasons for that. Giving software free access to your life savings is very different from asking it to order paper towels.
Still, the situation illustrates the broader issue. Friction can protect a business, whether that friction exists for good reasons or not. As agents become better at safely handling tedious financial tasks, some of that protection could weaken.
When figuring out the price is the moat
Sometimes consumers don’t stick with a company because switching is hard. They stick because comparing the alternatives is hard.
Insurance is a good example. Comparing two policies means looking beyond the headline premium to deductibles, coverage limits, exclusions and other details. Telecom plans can involve different data limits, bundles, promotional prices and fees.
A person may eventually decide the potential savings aren’t worth the effort.
Software doesn’t get frustrated after the sixth page of fine print.
This is another area where businesses are already preparing for change. Kin Insurance recently said it would welcome AI shopping agents rather than block them and has since built tools designed to make it easier for agents to obtain home insurance quotes. Insurance marketplace Insurify has also launched an AI Agent Gateway that allows agents to gather and compare real-time quotes while preserving details such as deductibles, limits and disclosures. (Kin / PR Newswire)
That may offer a preview of how companies adapt.
If AI makes comparison easier, businesses can try to fight the change—or make themselves easier for agents to shop. In some industries, being the company that gives an agent the clearest information and easiest path to a transaction could eventually become an advantage of its own.
When forgetting to cancel is the moat
Subscriptions may be the easiest behavioral moat to understand.
Many of us have looked at a credit card statement and found something we forgot we were still paying for. Sometimes it is a streaming service. Sometimes it is an app. Sometimes it is the gym membership that seemed like a great idea in January.
There is good evidence that this matters to businesses. A 2025 study published in the American Economic Review looked at subscription companies and found cancellations rose sharply when replacement credit cards forced customers to make an active renewal decision. The researchers estimated that cancellation friction roughly doubled seller revenues on average, holding the starting subscriber base constant. (American Economic Review)
An AI agent could make that kind of inattention much less common. It could regularly scan recurring charges, compare them with usage and ask whether you still want to keep paying.
That concern helped send Planet Fitness shares sharply lower following the launch of Muse, as investors considered what easier cancellation might mean for businesses that rely on recurring memberships. (StreetInsider)
Netflix is a more interesting case.
There may be some customers who keep Netflix because canceling is one more thing they haven’t gotten around to doing. Agentic AI could put pressure on that part of the business.
But Netflix also has something much harder for an agent to eliminate: content people actually want to watch. If a consumer wants a show that is only available on Netflix, an AI agent cannot shop around and recreate it somewhere else.
That is the difference between a behavioral moat and a real one.
When settling for good enough is the moat
Even when prices are relatively clear and switching is easy, comparison takes time.
Most people do not check every travel site each morning to see whether another hotel has dropped its price. We don’t constantly search every retailer to make sure something we bought last week is still the best deal. At some point, we stop looking and buy.
AI agents could keep looking.
Travel companies are already building around this idea. Priceline’s Penny can compare destinations, prices and availability and move from a broad travel request toward an actual booking within one conversation. (Priceline)
But perhaps the most interesting response is happening in retail, where Walmart and Amazon are taking very different approaches.
Walmart recently introduced an Agent Gateway designed to allow outside AI agents to connect directly with its systems, search products and make purchases for customers. Shoppers will be able to control what their agents can buy and set limits on their activity. (Walmart)
Amazon, meanwhile, has blocked Muse from shopping on its platform, citing concerns including security and unauthorized access. Amazon is also developing its own AI shopping tools rather than simply opening its ecosystem to outside agents. (Axios)
Both approaches may prove reasonable. This technology is new, and companies have legitimate reasons to be careful about letting outside software interact with customer accounts.
But the contrast between the world’s two largest retailers shows how important the issue could become.
Today, a consumer might go directly to Amazon or Walmart to shop for a television. Tomorrow, that person might simply tell an agent, “Find me the best 55-inch television under $700.”
If the agent does the searching, comparing and buying, the retailer may no longer own the first interaction with the customer. That could eventually change the value of advertising, loyalty programs and simply being the default place where consumers begin their search.
Companies including Mastercard are already building infrastructure around this possibility, helping merchants make products available to AI agents while maintaining control over pricing, fulfillment and transactions. (Mastercard)
What AI can’t shop away
None of this means traditional business moats are disappearing.
If anything, agentic AI may make it easier to tell which ones were real.
A patented product or valuable piece of intellectual property cannot simply be compared away. Neither can truly scarce physical assets. A competitor cannot instantly create another well-located piece of infrastructure, another ski mountain or another casino license in a market where those licenses are limited.
The same can apply to real-world experiences. AI may find a cheaper vacation, but it cannot recreate the exact cruise, resort or experience a customer wants.
Those businesses still have to compete. But their advantage comes from offering something difficult to duplicate, rather than relying primarily on consumers being too busy or frustrated to find an alternative.
That distinction could become increasingly important as AI agents get better.
For decades, some companies have benefited not just because they offered the best product, but because consumers were imperfect shoppers. We didn’t have enough time, information or patience to constantly make the optimal choice.
AI may not eliminate those imperfections overnight. Consumers still need to trust agents with their money and personal information, and companies will continue putting limits around what outside software can do.
But if agents begin taking more of the work out of shopping, switching and canceling, one question may become increasingly important for businesses across the economy:
How much of your moat comes from being genuinely better—and how much comes from customers not bothering to leave?
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune. Meristead Wealth is a Registered Investment Advisor with the United States Securities and Exchange Commission (SEC File No. 801-107061). This commentary is for informational purposes only and should not be construed as investment advice or a recommendation to buy or sell any security. As of June 30, 2026, as reported in their Form 13F filed August 14, 2026, their portfolio strategies held positions in Planet Fitness (PLNT), Meta Platforms (META), and Amazon (AMZN), which are mentioned in this article. Securities identified do not represent all securities purchased, sold, or held for the benefit of Meristead clients or by the employees of Meristead. Readers should not assume that investments in the securities discussed were or will be profitable. Views, examples, and holdings are subject to change without notice.
The post Which business moats will still matter in the age of AI? appeared first on Fortune.
