Software

The software industry designs and sells the programs, platforms, and cloud services that run nearly every other part of the modern economy, and it has spent the past several years going through arguably its biggest strategic shift since the move from on-premises software to the cloud: the emergence of generative AI as both a new category of product and a fundamental change in how software itself gets built. Unlike industries built on physical goods, software's economics are defined by extremely high fixed costs to build a product and extremely low marginal cost to distribute it to an additional customer — a combination that rewards scale and platform dominance more than almost any other sector.

The industry today spans several overlapping business models: traditional licensed software (increasingly rare as a standalone model), the now-dominant software-as-a-service (SaaS) subscription model that companies from Microsoft to smaller specialized vendors rely on, cloud infrastructure platforms (AWS, Azure, Google Cloud) that other software companies build on top of, and a fast-growing category of AI-native products — companies whose entire value proposition is built around large language models and generative AI rather than AI being a feature bolted onto an existing product. That last category has driven an enormous wave of new company formation and venture investment, alongside genuine anxiety among incumbent software companies that AI-native competitors, or AI itself, could shrink the market for traditional software categories where AI can now perform tasks that used to require dedicated tools.

Enterprise adoption of generative AI specifically has moved from experimentation to genuine deployment at a rapid pace: an estimated 65% of organizations now use generative AI in at least one business function, roughly double the adoption rate of just ten months earlier, and more than 80% of enterprises are expected to have deployed generative-AI-enabled applications by the end of 2026. Enterprise spending on generative AI reached an estimated $37 billion in 2025, more than triple 2024's roughly $11.5 billion — a growth rate that has made AI tooling one of the fastest-adopted enterprise technology categories in recent memory, comparable to or exceeding the early adoption curves of cloud computing and mobile enterprise software in prior cycles.

A defining current tension in the software industry is exactly how much of software development itself gets automated by AI coding tools, and what that means for the industry's own workforce and cost structure. AI coding assistants have moved from simple autocomplete to genuinely capable "agentic" tools that can write, test, and even deploy meaningful chunks of code with limited human oversight, raising real questions about how software companies' own engineering costs — historically their largest expense — evolve over the coming years, and whether that efficiency gain accrues mostly to incumbent software vendors or to whichever companies most successfully deploy AI-assisted development to out-build competitors faster.

Competitively, the software industry remains defined by network effects and switching costs more than by any single technology: once a company embeds a piece of software (an ERP system, a CRM, a cloud platform) deeply into its operations, the cost and disruption of switching to a competitor — even a technically superior one — is often high enough to entrench incumbents for years. That dynamic explains why the biggest platform companies (Microsoft, the major cloud providers, and a handful of dominant enterprise software vendors) have generally been able to add AI capabilities to their existing products and defend market share, even as a wave of AI-native startups compete hard for the parts of the market not already locked into an incumbent platform.