How exposed are software stocks to AI tools? We put vibe-coding to the test

CNBC | February 05, 2026 at 09:07 PM UTC
Bearish 79% Confidence Unanimous Agreement
Read Original Article

Key Points

  • Non-developers created a working Monday.com clone with AI integration in under one hour for just $5-$15 in compute credits, suggesting easier-to-replicate software may face significant competitive pressure
  • Most exposed companies include project management and workflow tools like Atlassian, Adobe, HubSpot, Zendesk, and Smartsheet that aren't core to business operations
  • Systems of record like Salesforce and cybersecurity stocks like CrowdStrike and Palo Alto Networks are considered safer due to network effects and enterprise data anchoring that make them harder to replicate

AI Summary

Summary: AI Coding Tools Pose Threat to Software Stocks

CNBC conducted a practical test to assess AI's threat to software companies by using Claude Code, an AI coding tool, to replicate Monday.com's project management platform. Two non-developers successfully created a functioning prototype in under an hour, costing only $5-$15 in compute credits.

Key Findings

Most Vulnerable Companies:

Silicon Valley insiders identify software tools that "sit on top of the work" as most exposed, including:

  • Atlassian
  • Adobe
  • HubSpot
  • Zendesk
  • Smartsheet
  • Monday.com (market cap: $5 billion)

These are considered "nice-to-have" rather than core business tools, making them easier to replicate with AI.

More Protected Companies:

  • Cybersecurity stocks (Crowdstrike, Palo Alto Networks): Network effects and critical security functions make them harder to disrupt
  • Systems of record (Salesforce): Anchored by enterprise data and deeply embedded in business operations, making replication via weekend coding projects impractical

Market Implications

Software, legal services, and video game stocks have experienced significant sell-offs recently due to AI disruption fears. The experiment demonstrates these concerns have merit—AI coding tools enable non-technical users to build functioning applications quickly and cheaply using plain English commands.

As data center infrastructure expands, compute costs are expected to decline further, potentially accelerating this disruption.

The wholesale software sector sell-off may present opportunities for investors to distinguish between essential "need-to-have" platforms with strong moats versus vulnerable "nice-to-have" tools facing genuine AI displacement risk.

Model Analysis Breakdown

Model Sentiment Confidence
GPT-5-mini Bearish 80%
Claude 4.5 Haiku Bearish 72%
Gemini 2.5 Flash Bearish 85%
Consensus Bearish 79%