The AI-fueled software meltdown is overblown

The software sector has taken a massive beating on Wall Street over the past few months on fears that AI is destined to roll over the segment.

But the sell-off turned into a meltdown last week after Anthropic (ANTH.PVT) released an update to its Claude Cowork tool that included plugins to perform common tasks related to legal, marketing, finance, data, and sales. OpenAI (OPAI.PVT) then followed with its own new release, and Anthropic launched yet another.

ServiceNow (NOW) stock is off more than 22% since Jan. 29, while Thomson Reuters (TRI) is down more than 26% as of Wednesday. Intuit (INTU) shares fell more than 26%, and Snowflake (SNOW) shed 18%. Salesforce (CRM) dropped more than 20%.

The thinking is that Anthropic, OpenAI, and other AI companies will either build their own software to rival offerings from existing software firms or allow businesses to easily develop custom in-house software. Either scenario could spell trouble for software companies.

But according to some analysts, the widespread panic on Wall Street is an overreaction. And instead of replacing all existing software companies, AI will likely allow many incumbents to augment their services to better serve their customers.

But that doesn’t mean all software companies are safe.

“I do think it’s overblown for a large part of the software industry,” William Blair analyst Jason Ader told Yahoo Finance. “You have a baby-with-a-bath-water situation right now, or, you know, software indexes are just getting sold, and obviously no one necessarily pays attention to who’s in the index.”

He added, “I think we have to be a little more discerning, discriminating on which software companies are more at risk versus others.”

Read more: How to protect your portfolio from an AI bubble

The AI industry has made some major advancements since OpenAI debuted ChatGPT in 2022. And while it has certainly changed the way many of us browse the web, draft emails, and catch up on meetings we’ve missed, it’s too early to crown AI companies the new enterprise software kings.

Still, even the hint of a challenge to the current order of software companies is enough to rattle Wall Street. And that uncertainty has helped fuel the broader software sell-off and sent investors fleeing to perceived safe havens.

“If there’s more unknowns around that long-term outlook, then I think you’re going to have investors value the companies at a lower level and certainly think about rotation into other sectors where there’s more certainty, like memory and chips and data center infrastructure and power and utilities and HVAC systems,” William Blair’s Ader explained.

But there are several roadblocks that could keep AI companies from crushing existing software firms underfoot.

According to Morgan Stanley’s Keith Weiss, businesses are unlikely to allocate their IT resources to have open-source AI models write and maintain their own custom software unless it provides a key differentiator over the long term.

“The key point here is that initial development cost is only one part of the equation when organizations think about whether to automate business processes with their own custom coding or partnering with 3rd party vendors,” Weiss wrote in a research note.

OpenAI CEO Sam Altman attends an event to pitch AI for businesses in Tokyo, Japan February 3, 2025. REUTERS/Kim Kyung-Hoon — OpenAI CEO Sam Altman pitches AI for businesses in Tokyo, Japan, on Feb. 3, 2025. (Reuters/Kim Kyung-Hoon) · REUTERS / Reuters

“We would note open-source software, which is free of licensing costs (essentially free development), has been available for customers to build their own applications for over 2 decades. Despite this, the market of third-party software has flourished in that time,” Weiss added.

Ader offered a similar take, saying that it doesn’t make sense for a company to “vibe code” a new customer relationship management system or payroll app using something like Claude Code.

“If you have something that works pretty well, as long as it’s not overly expensive, you’re going to want to focus on your core competencies,” he added.

As for whether model builders will steal market share from existing software companies, Weiss wrote that large language models (LLMs) are unlikely to “ever be as effective and efficient of a data warehouse or messaging server as the existing solutions purpose built for those capabilities.”

Instead, he explained, various AI features will likely end up built into existing software, expanding its capabilities and value for customers.

There’s one other problem with AI companies being the new software businesses: data governance.

Organizations are generally protective of their proprietary data and may not be comfortable opening their troves of information to AI models or AI companies, rather than to companies they’ve worked with for years.

None of this, however, means that all software firms will be spared. Some may not be able to keep up with the constantly evolving AI landscape and will inevitably fall behind. But those that can keep pace with AI will likely continue on.