Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →No: the $1 trillion figure does not mean SaaS companies lost $1 trillion in sales. It refers to a reported decline in the market value of the S&P 500 software and services index during an early-2026 selloff. The more useful conclusion is that AI is changing what software does, how vendors compete, and how buyers judge value—not that the subscription software model has disappeared.
What does the $1 trillion SaaS selloff mean?
A February 2026 Reuters report carried by Investing.com said the S&P 500 software and services index had shed about $1 trillion in market value since January 28 during a selloff. That is a change in what investors valued the index’s companies at over a particular period. It is not a tally of lost SaaS revenue, customer spending, or proven economic damage.
As an Amazon Associate I earn from qualifying purchases.
Investors were pricing in the possibility that generative AI and agentic systems could disrupt software businesses: if a new tool can perform work that once required several applications, buyers might switch vendors, pay less for existing products, or expect more capability for the same price. That is a risk being evaluated by markets, not proof that those outcomes have already happened across the industry.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesKeep the big AI and software numbers separate
Large figures in AI coverage often describe different things: stock-market value, forecast spending, modeled economic potential, or surveyed company outcomes. They cannot be added together or treated as the size of the SaaS market.
| Figure | What it measures | How to interpret it |
|---|---|---|
| About $1 trillion | Reported market-value decline in the S&P 500 software and services index since January 28 during the early-2026 selloff, according to Reuters in February 2026. | An investor-valuation change, not lost software revenue. |
| $2.670 trillion in 2026; $3.637 trillion in 2027 | Gartner’s September 2026 forecasts for worldwide AI spending. Gartner separately forecast $461.637 billion for AI software in 2026. | Forecast spending across AI categories, not realized totals or SaaS revenue. The overall forecast includes infrastructure and other markets beyond software. |
| $2.6 trillion–$4.4 trillion per year | McKinsey’s 2023 estimate of potential annual economic value across 63 generative-AI use cases. | Modeled potential if use cases are adopted across the economy, using the structure of the global economy in 2022; it is neither sales nor current spending. |
| $175 billion–$250 billion by 2027 | McKinsey’s 2024 forecast of possible generative-AI spending. | A dated forecast, not a reported 2027 result. |
| $643.860 billion in 2025 | Gartner’s March 2025 forecast for worldwide generative-AI spending that year; Gartner forecast 80% would go to hardware. | A forecast published in 2025, not a later measured result—and a reminder that AI spending totals include more than SaaS applications. |
The different definitions matter. For example, Gartner’s $2.670 trillion 2026 AI-spending estimate covers a wider set of markets than its $461.637 billion AI-software estimate. McKinsey’s potential-value range is a model of economic benefit, not a forecast of what software vendors will earn.
Is SaaS dying, or is AI changing the product?
The evidence points to change rather than a single, industry-wide ending. SaaS vendors are adding AI to products customers already use, while AI-native tools and agentic workflows may take over parts of some applications. The pressure is real, but its consequences depend on the software category, the workflow, and whether a new product can do more than generate a plausible response.
Gartner’s September 16, 2026 release says vendors are “rapidly embedding agentic AI within their existing products to maintain relevance in the market and defend against new cross-functional agents.” This describes a competitive response: incumbents are trying to make existing products more capable as new tools challenge the boundaries between applications.
Free tools Windows power users keep installed
One-click scans. No signup required.
McKinsey’s 2025 paper by Mohit Khanna and Nimish Mittal, with Ashley Wu and Marc Castillo, describes the deeper shift: “AI is transforming software from a tool that enables work into a platform that actively performs and orchestrates work.” A product that takes action may change the user experience and pricing logic, but it still has to fit into the business process it is meant to improve.
How embedded AI and AI-native workflows differ
For a buyer, the practical comparison is not simply “old software versus AI.” It is whether an AI feature inside an existing application or a newer AI-native workflow can reliably complete the work, connect to the right business systems, and meet organizational requirements.
| Question | AI embedded in incumbent SaaS | AI-native product or agentic workflow |
|---|---|---|
| What work does it do? | May add assistance or automation within an application and its existing workflow. | May combine or automate steps across a workflow; verify that it completes the process rather than only producing an answer. |
| What data and controls does it use? | Assess the data access, integrations, permissions, and audit trails available in the existing environment. | Check how it connects to business data and systems, respects permissions, and records actions. |
| What does adoption require? | Assess implementation effort and whether users need new training to use the added capability. | Include migration, integration, implementation, and retraining in the switching cost. |
| What risks need review? | Review governance, security, data sovereignty, vendor dependence, and any AI-specific charges. | Apply the same reviews, including whether adding a new provider creates another dependency or governance boundary. |
| How should value be tested? | Measure the change against the same workflow outcome you would expect from a replacement. | Set a baseline and test an outcome such as handling time, quality, throughput, or a financial result. |
Gartner reports that enterprises are using simpler embedded features from incumbent providers and identifies lock-in, data sovereignty, and runaway costs as concerns that have not deterred adoption in general. That pattern is not a guarantee that a particular feature is suitable or that every organization accepts those risks. An agent also does not automatically remove the need for systems of record, integrations, permissions, compliance processes, or accountable human review.
Rank #4
Why AI can create growth and churn at once
AI can expand demand for software while making individual vendors easier to replace. McKinsey’s 2024 analysis estimated that generative AI might add two to six percentage points to software-sector growth, while accelerating vendor switching by five to ten percentage points and increasing total churn by one to three percentage points. These are analysis-based estimates, not measured outcomes or guarantees about what happened by 2027.
The tension is straightforward: more automation can create new software needs, but a product that fails to deliver useful results may lose customers faster when alternatives can perform part of its job. McKinsey’s same analysis cautions that long-term outcomes are difficult to predict given the technology’s evolution and the range of ways it could be applied.
Generative-AI use cases in McKinsey’s 2023 economic analysis include customer interactions, marketing-content generation, and drafting code from natural-language prompts. Customer operations, marketing and sales, software engineering, and research and development together represented about 75% of the modeled potential value in that analysis. Those are areas of potential, not evidence that every deployment has already delivered savings or improved performance.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What enterprise productivity evidence does—and does not—show
In a 2025 survey, McKinsey reported that 46% of surveyed companies said they had achieved scaled AI productivity impact or financial impact, compared with 33% one year earlier. The survey describes large companies and samples in North America and Europe; the figure should not be read as a global adoption rate or as proof that every AI project generated a financial return.
For buyers, the implication is to measure an implementation against a specific baseline rather than assume that adding AI produces value. Track an outcome that matters to the workflow—such as handling time, quality, throughput, or financial performance—and evaluate it alongside implementation and operating costs.
How to evaluate an AI feature or replacement
- Define the workflow. Specify the task, users, systems involved, and what counts as completion. Distinguish a tool that drafts a response from one that safely completes the full process.
- Check data access and permissions. Confirm what business data the tool can use, which integrations it needs, how access controls are enforced, and whether actions can be audited.
- Estimate the full change cost. Include implementation, migration, integration, retraining, and any costs of retaining or replacing connected applications.
- Review governance and exposure. Assess security, data sovereignty, compliance needs, vendor dependence, and responsibility for reviewing consequential actions.
- Make AI costs visible. Ask how usage is tracked, how charges vary with use, and what controls help prevent unexpected costs. Gartner identifies cost management and usage tracking as relevant enterprise concerns.
- Run a bounded outcome test. Compare results with a baseline using a defined measure—such as time, quality, throughput, or financial impact—and decide in advance what evidence would justify broader adoption.
What SaaS vendors have to change
The selloff and the AI forecasts point to a strategic test for vendors: adding a feature is not the same as making a product more valuable. Vendors need to show that AI improves a meaningful workflow, works with the data and controls customers require, and has costs customers can understand. If AI makes a product easier to replace or shifts value from individual applications to cross-application workflows, retention and differentiation become more important.
The market is still uncertain. Gartner’s September 2026 forecast also describes a $1.2 trillion opportunity in AI services by 2030, a separate forecast from the stock-market decline and from software revenue. These projections signal the scale of expected investment and implementation work, not a guaranteed outcome for any vendor. SaaS is changing because both buyers and providers are testing where AI belongs in business software—and which products can turn that capability into dependable results.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
