The Selloff
Australian software stocks lost $8.8 billion in five days as investors bet that AI agents could replicate core SaaS functionality. Atlassian dropped 7.7% on February 4 despite reporting 23% revenue growth to $US1.6 billion. The company's shares are down 50% in 12 months.
REA Group shed more than $7 billion in market value over the same period. Web Travel Group collapsed 29.5%, leading the rout across ASX tech names.
What Changed
The panic followed announcements from Anthropic and OpenAI showcasing AI agents capable of complex workflows. US markets lost roughly $300 billion across software and data stocks overnight February 3-4. Application software now trades at 20x forward earnings, below historical averages.
Alphabet's plan to double AI capital expenditure in 2026 added fuel: if the platform providers are spending this much, the thinking goes, they're building to replace, not augment, existing software layers.
The Defense
"I'm convinced AI is great for Atlassian. Others think software is dead," CEO Mike Cannon-Brookes told investors after earnings. He pointed to accelerated product development using AI tools.
REA CEO Cameron McIntyre emphasized "the speed at which we're now developing new capabilities with AI" as evidence the company can adapt faster than competitors.
The pitch: incumbents with deep customer relationships and proprietary data can integrate AI better than new entrants. History suggests this works until it doesn't.
The Context That Matters
This isn't just sentiment. US job openings hit a 63-month low in January. Tech layoffs reached levels not seen since 2009. Enterprise software buyers are reconsidering seat-based licensing as AI potentially reduces headcount needs.
Vendors are shifting to usage-based pricing models. The question for CTOs: which platforms prove they can capture AI value, and which get automated away?
What to Watch
Valuations have reset significantly. WiseTech and Xero are down over 50% from peaks despite strong fundamentals. Some analysts see buying opportunities if companies demonstrate credible AI roadmaps.
The alternative view: 2026 becomes a sorting mechanism. Best-in-class operators with sticky enterprise customers survive. Everyone else faces margin compression as AI commoditizes their differentiators.
The contagion is already spreading beyond software. Private credit stocks, cryptocurrency markets, and even uranium producers saw selling as investors reassessed AI infrastructure build-out assumptions.
For enterprise tech leaders, the message is clear: show your AI adaptation strategy or watch your vendor's stock price (and potentially their viability) decline.