As news about generative artificial intelligence (GenAI) continually splashes across social media feeds, including how  ChatGPT 4o can help you play Rock, Paper, Scissors with a friend, breathtaking pronouncements about GenAI’s “disruptive” impact aren’t hard to find.

It turns out that it doesn’t make much sense to talk about GenAI as being “disruptive” in and of itself.

Can it be part of a disruptive innovation? You bet.

But much more important than just the AI technology in determining whether something is disruptive is the business model in which the AI is used—and its competitive impact on existing products and services in different markets.

What is Disruptive Innovation?

A disruptive innovation is not a breakthrough innovation. It’s one in which a relatively simple product or service initially takes root among people who don’t have the expertise or money to use the dominant products or services in a given market. From there, it improves and tackles more and more complicated challenges until it begins to displace market-leading products and services.

Importantly, disruption is a theory of competition. Something can’t be simply and absolutely “disruptive.” It must be disruptive relative to something else.

Also important is that although there must be a technology enabler that allows a disruptive innovation to improve and go “up-market” without replicating the expense, complication, and inconvenience of the existing products and services in a sector, a new technology is rarely, if ever, inherently “disruptive.” New technologies that can be used to spur a disruption can also be used as what we call “sustaining” innovations. Sustaining innovations are absolutely critical because they’re how existing products and services improve over time.

What matters is where the technology is used relative to a product or service’s business model. Is it allowing a low-cost, convenient, or simple value proposition to take root? Or are we deploying the technology in an existing business model to improve what we’re already offering?

Neither one is wrong. They just have different impacts on the market. One sustains the current trajectory of improvement. The other disrupts it.

AI as sustaining innovation

What’s clear so far is that the dominant technology firms—Amazon, Apple, Google, Meta, and Microsoft—are largely (and predictably) using AI as a sustaining innovation. For the most part, it’s helping them improve their current offerings.

Ben Thompson has written at length about this, so I won’t repeat his points. But for these firms, the biggest risk seems to be for Google—namely that AI upends its search-based business model through an offering that, yes, isn’t as good initially as Google Search but is also something that Google can’t monetize and that improves reliably and perhaps rapidly.

The question is about how and where the technology is used. Is it improving a current business, or is it placed in a new business serving nonconsumers that will then improve over time to tackle more complex use cases?

AI as disruptive innovation

The obvious next statement is that OpenAI may prove disruptive to particular products through ChatGPT. In education, tutoring and Chegg are obvious examples. But incumbent firms are also likely to use AI to improve their own offerings as a sustaining innovation.

Perhaps the most important thing to keep an eye on is where AI can radically transform industries through whole new business models that create new value networks.

In education, for example, imagine a microschool that leverages AI to dramatically lower its costs and create something far more scalable and affordable that doesn’t degrade as it expands. Will Alpha schools settle into that slot through its 2-hour learning platform? What about Khan Lab School and Khan World School through Khanmigo? 

Or imagine a new higher education offering that uses AI to create a significantly more affordable, personalized, and contextual learning experience but also uses the technology to lower the cost of acquisition of new students dramatically. Will a new venture being incubated by Southern New Hampshire University (SNHU)—and headed by Paul LeBlanc—step into this lane?

Or think about the tablet-based learning innovation that Imagine Worldwide has helped pioneer with onebillion in Africa to allow children throughout Africa to become numerate and literate at less than $7 per student per year. If they add AI into the software, might new “schooling” models be built that dramatically scale excellent and radically affordable learning, which can then be altered to meet different local contexts and priorities?

Either way, what’s clear is that AI isn’t inherently disruptive or sustaining. It’s how it’s used. It’s the business in which it is used.

The bottom line is that the new business model must:

  • Initially target people who are nonconsumers or overserved by existing products and services
  • Be not as good as existing products and services as judged by the historical measures of performance
  • Be simpler to use, more convenient, or more affordable. The use of the technology enabler must allow the new value proposition to move up-market without replicating the existing costs and structures of the existing products and services
  • Not alarm the incumbent business models. Incumbents must be motivated to ignore the new disruptive uses of the technology because they don’t initially feel threatened—nor do they see a way to serve these new markets profitably.

If a new business model powered by AI passes those tests relative to a specific set of products and services, then something disruptive might be taking root.


  • Michael B. Horn
    Michael B. Horn

    Michael B. Horn is Co-Founder, Distinguished Fellow, and Chairman at the Christensen Institute.