Software’s AI Reality Check: Why US Software Stocks Just Got Smacked

US software stocks have had a rough start to 2026.

Big software names like Microsoft, Adobe, Salesforce, and ServiceNow have dropped sharply this year, and some are far below their all-time highs.

The iShares software ETF, IGV, is down roughly in the low 20s percent year to date, even after a bounce. At the same time, software-focused indices are in or near bear market territory, while the broader market is roughly flat or up.

This matters even if you don’t own these stocks directly, because software and cloud businesses are a big chunk of tech funds, growth funds, and many broad market funds.

And this week, investors got a fresh reason to worry: new automation tools from Anthropic, built around Claude, that put pressure right on some of software’s most profitable use cases.

So if your portfolio has felt a bit shakier than some market headlines suggest, software could be a reason why. Let’s unpack what actually spooked investors, and what it could mean if you’re investing for the long run.

Why Software Is Suddenly Out of Favour

The simple version is this: investors are suddenly worried that AI could disrupt the classic software business model.

For years, many software tools felt hard to replace. If your team used Adobe for design, Salesforce for sales, ServiceNow for workflows, or Duolingo for language learning, those products were the default choices.

Investors were willing to pay high prices for these companies because they expected years of steady growth from millions of paying users.

Now the fear is much more direct: AI tools might start doing the same jobs, either as stand‑alone apps like chatbots and image generators, or as custom tools that companies build for themselves instead of paying for a traditional subscription.

Recent earnings and outlooks from big software firms, including names like ServiceNow and SAP, have fed a new fear that growth could slow, pricing power could weaken, and competition could rise as AI-focused tools get better and cheaper.

1. Third‑party AI tools could replace legacy software altogether - for example, creators using chatbots or image models instead of Adobe.
2. In‑house AI tools could replace parts of enterprise software - for example, someone inside a big company building their own project‑management bot instead of paying for another seat on a SaaS platform.

To stay competitive, many older software companies now need to add powerful AI features into their own products and, in some cases, accept lower prices. That can help users, but it also risks putting pressure on earnings if companies end up selling fewer “seats” or need fewer humans to get the same amount of work done.

So this sell-off is not just about what happened last quarter. It’s about expectations being reset. Markets are changing how much they are willing to pay today for software profits in an AI-shaped future.

Claude Plugins Made the Fear Feel Real

On February 3rd, the fear went from abstract to very concrete.

Investors woke up to a clear signal of how fast AI is moving into work that used to require dedicated software and specialised teams.

In a blog post, Anthropic announced new plugins that can customise Claude and automate tasks across legal work, sales, marketing, and data analysis.

That matters because those are areas where many software companies, and professional services providers, have traditionally made strong money.

Investors reacted fast.

The selling first hit firms seen as exposed to legal, data, and analytics services. It then spread to big enterprise software names like ServiceNow and Salesforce.

One trader described it as the “SaaSpocalypse”. SaaS is short for “software as a service,” which usually means subscription software you pay for monthly or yearly, instead of buying once.

In the US, the selloff was big enough that hundreds of billions of dollars in market value were wiped out in a single day, and the weakness then spread across Europe and Asia.

OpenClaw and the New Disruption Feeling

To understand why the fear feels so intense, it helps to look at what’s been grabbing attention in AI right now.

OpenClaw is an open-source AI agent. An AI agent is different from a chatbot - a chatbot talks, while an agent can do things.

OpenClaw can follow instructions, connect to other tools and services, and take actions on your behalf.

It was launched just weeks ago by Austrian developer Peter Steinberger, and it has already gone through name changes, including Clawdbot and Moltbot. Its rise has been fuelled by social media attention and the idea that AI agents could move beyond answering questions to actually completing work.

People have documented it automating tasks like managing emails and calendars, browsing the web, and interacting with online services. Early integrations have also been on messaging platforms, letting users control the agent through text commands.

The exciting part, especially for investors, is the idea that a tool like this can take a short instruction from a person and then handle a long list of follow‑up steps on its own - the kind of steps that used to be spread across different apps and teams.

For software investors, the point isn’t that OpenClaw will replace everything tomorrow. It’s that tools like this make disruption feel closer.

If AI agents can automate more tasks, or interact with online services directly, they could change how some software products are used, how they’re priced, and who captures the value.

The Market Mood Shift Inside AI

Big Tech and the biggest cloud players are spending huge sums on AI infrastructure. At the same time, investors are starting to ask tougher questions about who really benefits, and who gets squeezed.

The mood is moving from “you can’t afford to miss AI” to “show me how this turns into durable profits.”

In that kind of market, investors often rotate within the AI theme (meaning move their money from one part of the AI value chain to another). Some parts closer to the physical infrastructure of AI can hold up better than parts that rely on software pricing staying strong, even as new AI tools flood the market.

What This Means for Long-Term Investors

Here are the takeaways if you’re investing for the long run.

First, a sector being down 20% to 30% does not automatically mean “cheap” or “broken.” It often means expectations are being reset. If the market is less confident about future growth, it will pay less for those future profits.

Second, AI is both a threat and an opportunity for software. Some companies will use AI to make their products stronger and more valuable. Others may struggle if AI makes parts of their offering easier to replace.

Third, if you own broad funds, you probably already have software exposure. That’s not a problem. It just means you will feel these swings sometimes, even if you never bought a single software stock.

Fourth, you don’t need to bet on one company, or one AI agent. Headlines like OpenClaw can make it feel like everything changes overnight. But long-term investing is not about guessing the next viral tool. Diversification and time horizon matter more.

And finally, focus on business models, not buzz. A simple question goes a long way: how does this company make money, and could AI tools realistically replace what customers pay for today, or do they make the product even more essential?

In other words: in an AI world, “great software” still matters, but investors are no longer willing to pay for the story until the numbers prove it.