AI Isn’t an Economic Moat Killer, But It Will Disrupt Industries

Artificial intelligence will challenge companies whose strengths depend on workflow frictions, labor intensity, and application stickiness, while preserving—or even strengthening—businesses built around infrastructure, proprietary data, network effects, or specialized domain workflows.

Plenty of AI impact is baked into stock prices today, but we also expect more volatility in the near future as AI labs, eventually aiming to go public, continue to rapidly iterate their agentic offerings.

As the market narrative oscillates between AI-lab maximalism and incumbent advantages, we see opportunities to buy resilient businesses at reasonable prices after recent selloffs.

To address AI effects on company moats, we have reevaluated the moat ratings for 132 companies where we felt AI could be disruptive and therefore warrant deeper analysis.

Download Morningstar’s guide to moat ratings amid AI disruption.

Measuring the Impact of AI on Economic Moats

The Morningstar Economic Moat Rating summarizes the length of a company’s competitive advantages. An economic moat is a structural feature allowing a firm to generate excess profits over a long period. If Morningstar analysts believe that excess returns will persist for 20 years or more, that company earns a wide moat rating. Analysts expect a company with a narrow moat to remain competitive for at least 10 years.

We developed a proprietary scoring framework to systematically evaluate how AI affects the durability of moats. This system was primarily aimed at evaluating moats in software, and small adjustments were made to make it more applicable to other select industries such as IT services or financial services.

The framework is designed to capture both the threats AI poses to existing moats and the opportunities it creates for companies well-positioned in the AI value chain.

Key Dimensions for Thinking About Moats in an AI World

		7 Key Dimensions for Thinking About Moats in an AI World

Source: Morningstar. Data as of March 13, 2026. Download CSV.

AI Does Not Uniformly Destroy Moats—It Acts as a Sorting Mechanism

Of the 132 companies reviewed, 22 wide moats were downgraded (20 to narrow, two to none), 18 narrow moats were downgraded, and two narrow moats were upgraded to wide on account of their infrastructure layer positioning and network effects. AI does create risk, there’s no question about that. But are those risks equal across all companies? Are there no software firms with moats remaining? Absolutely not.

Payroll services, IT services, and enterprise software were groupings that felt the most pressure from moat downgrades, negative revaluations, and increases in uncertainty. This makes intuitive sense, given that AI disruption tends to hit hardest when companies monetize human labor, simple workflow automation, and seat-based software licenses.

Interestingly, out of the companies downgraded, many still have large existing user bases, popular products, and/or important customer relationships. In other words, it was not black and white, even for many of the downgraded companies.

That said, we do think AI decreases long-term visibility materially, and it can weaken the durability of existing advantages by making parts of the workflow layer easier to replicate, easier to automate, or less dependent on seat growth.

Where Companies Showed Resilience

We don’t think AI is a universal disruptor of all competitive advantages; instead, it’s a sorting mechanism. Most company moats remained unchanged, but the number of companies downgraded was significant.

A company’s competitive advantage was usually secure when it benefited from network effects, controlled infrastructure or proprietary data, had deep and complex ecosystems, operated within high regulatory barriers, or had unique domain logic.

However, if a company mainly profited from workflow inefficiencies or user habits at the application level, its advantage tended to be less reliable.

We found the most resilience in complex engineering software workflows, cybersecurity, financial infrastructure firms, and offerings with unique data and/or network effect assets. Half the companies with wide moat ratings after this exercise exhibit network effects. Other key sources of resilience included regulatory barriers to change or other switching cost barriers, and unique data assets.

		Design Software, Platforms, and Cyber Were More Resilient

Source: Morningstar. Data as of March 13, 2026. Download CSV.

How the Sources of a Company’s Moat Were Affected

Companies with network effects as a moat source saw the fewest relative number of downgrades. Network effects are not based on the product’s technology; they are based on the strength of the network. It stands to reason that even as technologies change, the network itself may still be hard to disrupt.

Classic examples here included payment networks, exchanges, and even travel networks like Booking Holdings BKNG. The core of Booking’s business is reliant on the strength of the travel network and the ability to aggregate the long tail of hotel supply, something that technology innovations like AI do not necessarily solve for.

Network effects are also central to the two upgrades that came out of this process, Cloudflare NET and CrowdStrike CRWD.

Cybersecurity solutions will be in more demand than ever as AI proliferates and opens up seemingly unlimited attack vectors—just imagine nefarious agents acting 24/7, limited only by the amount of computing power available. Companies like Cloudflare and CrowdStrike also have structural data and scale advantages they derive from their networks, which led to our moat upgrades.

		AI Moat Review Outcome Distribution

Source: Morningstar. Data as of March 13, 2026. Download CSV.

Switching costs were not as effective, and close to half the downgraded companies were classified as having switching-cost moats at one point. In the current AI-focused environment, switching costs—historically used to support software moat ratings—require careful reassessment, especially as AI models become integrated into fragmented, complex enterprise technology stacks.

AI could automate key technological processes that were historically sources of switching costs (like data transfers) and could hurt end-demand for certain software incumbents. The increase in uncertainty around what the future of software looks like in the AI world made it difficult for us to have confidence in return structures more than 10 years out for many firms, leading to the highest concentration of downgrades.

Given the disparity between switching costs and network effects, it’s likely that as workflows become more AI-driven, simple customer embedding (a precursor to switching costs) is likely to face pressure, but genuine network effects—where scale improves liquidity, relevance, content depth, telemetry, or ecosystem utility—are expected to increase in value.

		AI and Economic Moats: Which Stocks Are Most at Risk?

		Behind the scenes of Morningstar equity analysts’ review of the economic moats for 132 companies.
	





			26m 49s
		 Mar 10, 2026

Watch