The software sector's worst nightmare just got real. CNBC reporters with zero coding experience used Anthropic's Claude Code to build a functioning Monday.com replacement in under an hour for less than $15. The experiment provides the first empirical evidence behind the fears driving a 30% selloff across enterprise SaaS stocks, as investors scramble to separate which software companies face existential AI threats versus temporary disruption.
The AI disruption everyone's been warning about just moved from theory to reality. Two CNBC reporters with no coding background sat down with Anthropic's Claude Code and emerged an hour later with a functioning replacement for Monday.com, the $5 billion project management platform that's become a staple for teams worldwide.
The experiment wasn't planned as a takedown. CNBC's Deidre Bosa and reporter Jasmine Wu simply wanted to test whether the existential dread gripping software stocks had any basis in reality. "We didn't expect to get anywhere," Wu wrote in the published experiment. What they found should terrify every SaaS executive watching their stock price crater.
They started simple, instructing Claude to build a project management dashboard with multiple boards, team assignments and status dropdowns. The AI agent spit out a working prototype in minutes. Then they got ambitious, asking Claude to research Monday.com independently, identify its core features and recreate them. The AI complied, adding calendars and additional functionality without human intervention.
But the real moment came when they connected their clone to an email account. The AI transformed into a personalized project manager, surfacing a forgotten birthday party invitation, flagging unsigned waivers and adding travel booking reminders. It cost somewhere between $5 and $15 in compute credits, depending on iteration cycles. As hyperscalers build out more data centers, that cost trajectory points in only one direction.
This is vibe-coding in action. These AI tools let users with limited technical skills build functioning applications through plain English commands. Replit launched mobile app vibe-coding features in January, while observers have noted that American and Chinese AI coding agents are reaching impressive capability levels. The technology's maturation timeline just compressed from years to months.
Silicon Valley insiders are already drawing battle lines around which companies survive the shakeout. The most vulnerable targets are what they call tools that "sit on top of the work" - software that's useful but not mission-critical. Atlassian, Adobe, HubSpot, Zendesk and Smartsheet land squarely in the crosshairs. These are nice-to-have productivity layers that AI agents can potentially replicate with startling speed.
Cybersecurity stocks occupy safer ground. CrowdStrike and Palo Alto Networks benefit from network effects and security moats that no rational actor would try replicating with weekend coding sprints. The compliance burden and liability exposure alone create natural barriers.
Systems of record present a murkier picture. Salesforce anchors businesses with enterprise data and integration depth that's genuinely difficult to clone. But "difficult" isn't the same as impossible, and the Monday.com experiment proves that AI coding agents can tackle surprisingly complex application architecture when given clear direction.
The wholesale software sector selloff this year reflects this uncertainty. Investors are repricing decades-old assumptions about software moats and switching costs. If two journalists can build a credible Monday.com alternative in 60 minutes, what happens when motivated startups or enterprises apply the same tools to their specific pain points?
The compute economics accelerate the threat. At $5 to $15 per build, companies can prototype multiple custom solutions for less than a single Monday.com enterprise seat. As Anthropic and competitors scale inference infrastructure, those costs drop further while capabilities improve. The cost curve favors disruption.
Monday.com's market cap hovers around $5 billion, pricing in growth assumptions that looked reasonable six months ago. Now investors face uncomfortable questions about whether project management platforms constitute defensible businesses or temporary arbitrage opportunities waiting for AI tools to mature. The CNBC experiment suggests the answer leans toward the latter.
The test also exposes broader vulnerabilities across the SaaS landscape. Vertical software companies built on relatively straightforward database structures face similar risks. Marketing automation, HR management, expense tracking - any category where the core functionality revolves around data organization and workflow automation could theoretically be vibe-coded by determined users.
What the experiment doesn't answer is whether AI-built clones can match the polish, reliability and support infrastructure of established vendors. Monday.com employs hundreds of engineers continuously refining performance, security and user experience. A one-hour prototype won't include robust error handling, scalability or the accumulated learnings from millions of users.
But that gap is closing fast. AI coding agents improve with each model release, while the barrier to entry for software creation collapses toward zero. The question isn't whether vibe-coding will disrupt software companies - the CNBC test proves it already can. The question is which vendors adapt fast enough to stay relevant when their customers realize they might not need them anymore.
The Monday.com experiment transforms software disruption from abstract threat to documented reality. Investors now have empirical proof that AI vibe-coding tools can replicate multi-billion dollar platforms in under an hour for negligible cost. The selloff across software stocks isn't panic - it's repricing. Companies anchored by deep enterprise data integration and network effects have breathing room, but productivity tools sitting on top of work face genuine existential pressure. The line between need-to-have and nice-to-have software just became the most important valuation metric in tech.
