Angst is rising in the global software industry, as reflected by the S&P Software & Services Select Industry Index, down 21% year to date. Bellwether stocks are also slipping, such as Salesforce Inc. (ticker: CRM), down 28% in 2026, and Adobe Inc. (ADBE), down 23%.
The anxiety reached a fever pitch in early February, after Anthropic unveiled a new agentic artificial intelligence tool, Cowork, a non-coding version of Claude Code. The kicker: Cowork has been available to Pro subscribers in research preview since mid-January.
[Sign up for stock news with our Invested newsletter.]
The wider tech sector’s troubles didn’t start in February, nor in January. CRM shares are down about 45% in the past 12 months, and Adobe has slid more than 38% over the past year as software stocks have largely underperformed. The culprit seems to be an existential threat from artificial intelligence, and recency bias is not the main issue.
According to Credit Benchmark’s proprietary credit trend data, the software industry has experienced significant deterioration over the last four years. The data, which includes aggregated internal credit-risk views from more than 40 leading global financial institutions, paints a troubling picture for software companies.
Both the global and North American software industries show credit erosion since the December 2021 analytical baseline. Global software has 13.9% deterioriation as of December 2025, while North American software is at 4.2%. So, global software credit risk is roughly three times higher than in North America individually.
“Credit deterioration started meaningfully in 2023, which coincides almost exactly with when AI disruption fears emerged as a dominant narrative,” says David Carruthers, research advisor at Credit Benchmark. (ChatGPT launched in November 2022, with widespread adoption through 2023.)
“Despite a rally in 2024, credit risk has trended higher, peaking in late 2025 right as the equity market sell-off for software intensified,” says Carruthers. “Credit markets were pricing in these risks nearly two years before equity investors.”
A New AI Threat From the Legal Realm
Enter Anthropic’s new AI financial, legal and other industry tools on its Claude platform. On Jan. 30, Anthropic released 11 starter plug-ins to show what Cowork, an agentic, “no-code” version of the developer-focused Claude Code, can do in specific industries. A particular AI legal plug-in is built for the legal market and can aid law firms with paperwork-intensive tasks such as document review, risk flagging, NDA approvals, workflow automation and other necessary but often mundane legal chores.
Software investors can’t help but notice that Anthropic is pivoting from a third-party model tech supplier to an application-layer and workflow partner. That’s long been the role of legal and publishing software providers like RELX PLC (RELX), which owns LexisNexis, and of companies that provide information-based analytics, legal technologies and risk services to professional services firms.
The news from Anthropic further tilted the scales against software companies. The S&P Software & Services index promptly fell by about 12% by Feb. 5, resulting in several consecutive trading days of losses for the software industry. Clearly, strong forces are at work in the software space.
“The market isn’t overreacting to one AI tool; it’s repricing a business model that depended on customers having no alternatives,” says Ingrid Curtis, CEO of Sparq, an AI-native engineering solutions company.
Curtis notes that enterprise software-built valuations are based on high switching costs and annual price increases, which create challenges for software companies. “AI challenges whether those moats hold when you can build and customize software faster and cheaper,” she says. “Anthropic’s legal tool is just visible proof that specialized software functions can now be handled by adaptable AI systems.”
Previous technology cycles, such as software-as-a-service (SaaS) and cloud computing, changed how software was delivered, but the business model remained intact. Yet AI is different. “It shifts the value equation,” Curtis says. “When building or adapting software becomes materially faster and cheaper, buyers question what they’re paying for.” Organizations are already moving away from platforms that cost more each year and adapt more slowly.
“This isn’t fear, it’s repricing,” Curtis adds.
[READ: 7 Best Tech ETFs to Buy in 2026]
Taking the Long View on a Software Slump
That trend leaves software investors wondering about a structural shift in the enterprise software market. Or they may be weighing a familiar cycle in which new technology triggers fear before monetization realities catch up.
Which is it? Both, technology gurus say. That’s what makes trying to position agentic AI versus SaaS complex.
“The structural shift is real,” says A.K. Pradeep, CEO and founder of Sensori.ai, a neuroscience-powered generative AI company. “AI is fundamentally changing how software delivers value, moving from static tools to dynamic reasoning systems. This isn’t incremental improvement; it’s a categorical change in what software can do.”
However, enterprises don’t rip out working systems overnight. Instead, they retrofit and retool. Sensori.ai estimates that this scenario creates a three- to seven-year transition window during which software companies must evolve their products while maintaining existing revenue streams.
“The reality is that agentic protocols are being written up as we speak,” Pradeep says. “Agentic commerce is a bit of a rehash of modular smart code changes, monolithic SaaS. Now, we’ll see new hybrid models emerge.”
History has shown the pattern. “When SaaS emerged, on-premise software didn’t disappear; it transformed,” Pradeep says. “Enterprise customers took five to 10 years to fully migrate, and many successful software companies grew throughout that transition by offering hybrid models.”
3 Stocks That Should Benefit From the Software Shakeup
As any Wall Street philosopher will tell you, when one door starts closing, another one opens. Some technology companies will gain (or continue to gain) in stature as businesses question the merits of software. The outcome of the current sweeping changes is hard to predict, but here are some stocks to keep an eye on in this environment:
International Business Machines Corp. (IBM)
Pradeep advises tech investors to look beyond the obvious infrastructure plays such as Nvidia Corp. (NVDA) and Broadcom Inc. (AVGO). Instead, look to companies in IT services and systems that enable the actual enterprise transition. That’s where IBM enters the conversation.
“Every enterprise needs to retrofit existing systems and retool workflows around AI, and this isn’t a DIY project,” Pradeep says. “The consulting hours required to integrate AI into existing enterprise architecture will be enormous, potentially larger than cloud migration consulting.”
IBM should fit the bill. IBM is down 2.1% in 2026 as of Feb. 5, but it easily outperformed in the fourth quarter, leveraging strong software growth, hybrid cloud tech and the WatsonX AI platform, and improved cost efficiency. IBM is well positioned to capitalize on robust demand for hybrid cloud and AI, where it excels, particularly in data analytics and data security. Toss in a healthy 2.3% dividend yield, and Big Blue is a good bet for the post-AI software market.
Snowflake Inc. (SNOW)
This Bozeman, Montana, cloud data warehousing firm should benefit from a hybrid AI-software alliance. Its Data Cloud is an ecosystem where Snowflake customers, partners, data providers and data consumers can break down silos and derive value from data. That’s an area in great demand as AI models are only as good as the data they access. “Companies like Snowflake that help enterprises organize, govern and prepare data for AI applications become more valuable, not less,” Pradeep says.
SNOW shares should thrive as demand for data grows in the AI age. “Data platform integration with major AI models and enterprise AI deployment metrics is an area that bears watching,” Pradeep notes.
Snowflake’s recent lackluster stock performance (it’s down 28.6% year to date) will likely give investors pause, but the company is attracting new customers in droves, and its strong customer retention rate bodes well for future growth. Meanwhile, the company’s Snowflake Intelligence enterprise AI agent is taking off, with CEO Sridhar Ramaswamy calling it “the fastest ramp in product adoption” in the company’s history on a recent earnings call.
Amazon.com Inc. (AMZN)
As the AI revolution rocks the markets, it’s helpful to step back and recognize that AI announcements hit software stocks for mechanical and psychological reasons.
“Mechanically, long?duration, growth?sensitive software names are still very multiple?driven, and with the 10?year Treasury around 4% to 4.5%, any hint of slower growth or displacement gets aggressively discounted,” says Sutanto Widjaja, chief investment officer at Farther Institutional, a registered investment advisory firm.
Psychologically, investors have watched incremental AI dollars flow to Nvidia, Broadcom and hyperscalers, while software has been derated, Widjaja notes. “Consequently, each new AI product is read as a share loss for incumbents rather than incremental demand,” he says. “Many of these tools will ultimately be features inside platforms, but markets can’t model that path in real time.”
That’s where companies like Amazon, Nvidia and Broadcom earn a shout-out.
“If AI does ‘dominate’ the software layer, the biggest winners are the toll collectors on compute and cloud,” Widjaja says. He points out that Nvidia and Broadcom, for instance, monetize every training run and inference, making them efficient picks-and-shovels plays for the agentic AI era.
Yet Amazon, through AWS, is in a strong position as well for the run-up. “Amazon captures the infrastructure rent regardless of which application vendor wins,” Widjaja adds.
As AMZN is down 3.5% year to date, there should be value in acquiring shares at a decent price. Wall Street watchers seem to think so, with Mizuho analyst Lloyd Walmsley recently issuing an “outperform” call on the stock, with a $315 target price. AMZN closed at $222.69 on Feb. 5.
Takeaway
Market experts say software slippage due to AI is a temporary reevaluation, not a permanent decline. “We’re entering a retrofit-and-retool phase that will create massive opportunities for companies that enable enterprise AI transformation, particularly in consulting, integration and security,” Curtis says.
In that scenario, the winners won’t be companies replacing software; they’ll be companies helping software companies and their customers navigate this transition.
“The market is right to recognize AI’s transformative potential,” Curtis adds. “It’s wrong to assume transformation means elimination.”
More from U.S. News
10 of the Best Stocks to Buy for 2026
7 Best Semiconductor Stocks to Buy for 2026
8 Best Green Stocks and ETFs to Buy in 2025
Software Stocks Wilted After Claude Cowork’s Debut. But These 3 Stocks May Thrive originally appeared on usnews.com