News

AI Is Splitting Software Into Winners and Losers

The enterprise software market has divided into two camps this year — and which camp your employer sits in is starting to matter as much as your own skill set.

Software stocks lost more than $1 trillion in market value in early 2026, according to Forbes, with the iShares Expanded Tech-Software ETF falling over 20% as AI agents from major model providers moved into territory that legacy platforms once owned unchallenged. Salesforce dropped 26%. The disruption has been labeled the "SaaSpocalypse" by analysts, but underneath the headline selloff, a more nuanced story is taking shape: some software categories are accelerating while others face structural decline.

A Tale of Two Software Markets

The divide tracks a single variable — how deeply a company has embedded AI into its core architecture versus how much it has bolted AI onto an existing product. Sapphire Ventures' 2026 software report found that median growth for public software companies has fallen from 31% in 2021 to an expected 11% in 2026, a five-year slide. But within that average, Anthropic grew from $1 billion in annual recurring revenue at the start of 2025 to $14 billion by February 2026 — the kind of trajectory only available to companies built around AI from the ground up.

The companies analysts identify as clear winners share a common thread: they sit in the enabling layer, not the application layer. Snowflake, Datadog, Palantir, and CrowdStrike are all positioned as infrastructure or data-intelligence plays rather than seat-based productivity suites. The companies drawing caution are vertical SaaS and traditional per-seat applications — the kind of tools that automate human tasks rather than augment data infrastructure.

Gartner predicts that 40% of enterprise applications will include task-specific AI agents by year's end, up from less than 5% in 2025, according to Deloitte's 2026 software industry outlook. Incumbents that fail to make that transition face pricing pressure from a market that is actively shrinking its vendor count — the average number of SaaS applications per enterprise company dropped from 112 to 106 in 2025 alone, with 82% of companies saying they are reducing their supplier count.

What This Means for Job Seekers

Reviewing these signals, a pattern emerges that has real implications for where you plant your career flag. Roles supporting AI-native platforms and data infrastructure are on the growing side of this split. Roles inside traditional seat-based software companies — particularly those in the productivity, CRM, and collaboration categories that AI agents now challenge directly — carry more near-term risk.

That does not mean every legacy software job is precarious today. Fortune's reporting on the disruption found that mission-critical software tied to regulated or highly proprietary data remains defensible — companies are not rushing to put payroll or enterprise security through a general-purpose AI model. If you are at a software firm, the defensibility question worth asking is whether your employer's product handles data that customers are genuinely reluctant to hand to an outside model.

For those evaluating a move, our research suggests prioritizing platforms where AI is the foundation, not a feature. Data engineering, AI infrastructure, and platform-layer engineering roles are where the growth concentration is visible right now. Reviewing what AI skills command a salary premium in 2026 and which certifications are worth pursuing this fall are practical next steps for positioning yourself on the winning side of this split before the consolidation phase is complete.