For much of this year, the mood around software companies was getting harder to ignore.
Artificial intelligence was becoming better at writing code, automating tasks and building applications, raising an uncomfortable question for an industry that had spent decades selling software to businesses: what happens when customers can use AI to build some of that software themselves?
Investors reacted sharply.
The S&P 500 software and services index fell more than 26% between late January and its April low in a sell-off that became known as the “SaaSpocalypse.” But months later, the same sector is moving in the opposite direction. On Tuesday, the index rose 1.3% to its highest level since November 2025, according to Reuters.
So what changed?
AI Was Supposed to Be Software's Biggest Threat
The original fear was not difficult to understand.
If an AI system can write code, build an application or perform a task that once required a specialized software product, businesses could eventually decide they no longer need to pay for some of those products.
That possibility hit software companies particularly hard because their business models depend heavily on customers continuing to pay for digital tools year after year.
The anxiety became especially intense after AI companies began releasing systems capable of handling increasingly complicated work. In February, U.S. software and data-services stocks lost about $1 trillion in market value in a matter of days as investors worried that rapidly improving AI could fundamentally change the industry, according to reports
For a while, the question on Wall Street was whether traditional software had suddenly become yesterday's technology.
That prediction now looks much less certain.
The Software Rebound Is Sending a Different Signal
The latest numbers tell a very different story.
Reuters reports that the software and services index has gained about 5% this year, while expectations for the sector's 2026 earnings growth have climbed to 20.6%, up from 13.8% at the end of March. The sector also posted its strongest quarterly gain since the second quarter of 2020 during the July-to-September period.
That does not mean investors have decided AI is harmless.
Instead, they appear to be reassessing how quickly AI will disrupt established software companies and whether the technology might create as much opportunity as it creates pressure.
“AI has been more of an enabler for a lot of these software companies, more than a disruptor,” Adam Turnquist, chief cross-asset strategist at LPL Financial, told Reuters.
That sentence captures the shift.
AI may be changing the software business without necessarily destroying it.
Some Software Companies May Have a Better Defense Than Others
The recovery does not mean every software company is equally protected.
One of the most important defenses may be something AI cannot easily reproduce: proprietary data and years of deeply embedded business processes.
Earlier this year, Oracle executive Mike Sicilia pushed back against the idea that AI would simply wipe out software companies.
“You've all heard ... that new companies coding quickly using AI will spell the death of SaaS,” Sicilia told analysts, before arguing that the threat is different when software companies are rapidly adopting the same AI tools themselves.
Companies such as Oracle and Salesforce have spent years building systems around finance, customer relationships, supply chains, human resources and other business operations.
Replacing those systems is not necessarily as simple as asking an AI model to write a new application.
Businesses would still have to move data, change workflows, retrain employees and trust the replacement with critical operations.
That gives established software companies something AI startups may not have: years of relationships and infrastructure already embedded inside large organizations.
But Has the “SaaSpocalypse” Really Disappeared?
The fact that software stocks are recovering does not prove that AI will never disrupt traditional software. It only shows that the feared disruption has not happened as quickly as some investors expected.
Rebecca Wettemann, CEO of technology research firm Valoir, told Reuters that the “SaaSpocalypse” had not happened nearly as quickly as some people on Wall Street had expected. She said customers were increasingly moving beyond simply experimenting with AI.
That distinction could become important over the next two years.
Brian Mulberry, chief market strategist at Zacks Investment Management, told Reuters that a more serious test for software companies could arrive in the second half of 2027, when additional data-center capacity could make AI-powered coding more competitive with traditional software development.
In other words, Wall Street may not have abandoned the AI disruption story.
It may simply have decided that the ending is going to take longer to write.
What Happens Next
The next major test will come through earnings and actual customer behavior.
If software companies continue reporting strong demand while AI adoption accelerates, the argument that artificial intelligence can become an engine for the software industry will grow stronger. If AI coding and autonomous agents become capable of replacing large amounts of traditional software work, however, the disruption debate could return with even greater force.
For now, the feared software collapse has not arrived.
Instead, Wall Street is watching something more complicated unfold: the technology once expected to replace software is increasingly becoming part of the software business itself.
And that leaves investors with a different question than the one they were asking in April.
Not whether AI will change software.
But which software companies will change fast enough to benefit from it.
