SaaS rout deepens as AI fears hammer Wix and monday.com
Anthropic’s Claude Cowork triggers fresh doubts over subscription software, wiping billions from Israeli tech leaders.
The stock market is beginning to treat software companies as a legacy industry rather than the engine of digital growth. The release of Anthropic’s Claude Cowork has intensified fears that AI can perform much of the routine work for which businesses currently pay SaaS vendors.
Anthropic has presented Cowork, connected to its most advanced Claude language models, as an assistant capable of far more than writing code or answering questions. The tool is designed to analyze data, draft legal documents, prepare meetings, and handle a wide range of administrative computer tasks. The implication is unsettling for a sector built on the idea that every business process requires a dedicated software platform.
Investors reacted immediately. Shares of companies whose fortunes depend on workplace software fell sharply, with Israeli firms Wix and monday.com among the hardest hit. Wix dropped almost 7 percent in a single day and is now down more than 23% since the start of the year, 41% over the past six months, and 66% in the last year, leaving the company with a market capitalization of roughly $4.2 billion, its lowest level in more than two years. Monday.com fell nearly 10% on the day and has lost 29% this year and close to 60% in the past six months, reducing its value to about $5.2 billion, the lowest in more than three years.
The selloff reflects a broader reassessment of a model that once defined the technology industry. In 2011, venture capitalist Marc Andreessen declared that software was eating the world, arguing that nearly every business would be rebuilt around digital platforms. The thesis fueled a decade of investment in enterprise software startups and helped create hundreds of unicorns during the 2021 boom.
By 2026 the narrative has shifted. Software may have eaten the world, but AI is now eating software. The rise of natural language programming, so-called vibe coding, and increasingly capable AI agents is eroding the assumption that companies need large suites of specialized applications. Markets, often nervous but sometimes prescient, have begun to price in that possibility months before it becomes fully visible in corporate budgets.
The numbers tell the story. While the S&P 500 rose 17.6 percent in 2025, the SaaS index fell 6.5 percent. Household names such as Intuit, Atlassian, and HubSpot showed persistent weakness, and the first weeks of 2026 brought further double-digit declines. In Israel the reshuffling has been dramatic. Nice, monday.com, and Wix once ranked among the largest Israeli companies on Wall Street, but none of the pure enterprise software groups now sit in the top ten. Chip companies such as Tower Semiconductor and Nova have taken their place, reflecting a shift from software to hardware in the AI era.
The anxiety intensified after the viral release of Anthropic’s Claude Code, an agent capable of completing technical tasks that previously required dedicated developers. Stories circulated of a former Amazon executive who built a functional CRM system over a single weekend. Maor Shlomo, founder of Base44, tweeted that a customer had terminated a $350,000 Salesforce contract after recreating the solution with AI tools.
Such anecdotes have helped push valuation multiples lower. At the start of 2025 the median software company traded at more than seven times revenue; today the figure is below five. Hardware, long considered a slow and capital-intensive backwater, is enjoying the opposite momentum.
Yet the end of SaaS is not guaranteed. Large organizations have spent years and millions of dollars embedding ERP and CRM systems into their operations. Replacing them overnight is unrealistic. Many engineers warn that the flood of AI-generated code in 2026 will require extensive cleanup in 2027 as companies chase bugs introduced by automated development.
There is even a scenario in which traditional software usage increases. A marketing team that once sent a limited number of outbound messages may soon deploy an AI agent that triples activity, creating a new need for robust CRM systems to manage the resulting deluge.
