Anthropic's Claude Cowork launch revives investor fears about disruption that weighed on SaaS stocks in 2025; Morgan Stanley's SaaS index is down 15% YTD
The new year was supposed to bring opportunities for beaten-down software stocks. Instead, the group is off to its worst start in years.
Context & Ripple Effects
Concern over AI exposure had already become visible in Adobe’s share decline, with analysts tying its weakness to the perceived threat of AI-driven SaaS disruption Adobe’s earlier AI-disruption concerns. Claude Cowork broadens that debate from individual software categories to workplace-oriented AI capabilities.
The significance is less a verdict on immediate displacement than a renewed valuation test: investors are reassessing whether SaaS products retain differentiated workflows as AI tools reach a wider audience.
First-order effects
- SaaS equities face renewed selling pressure, reflected in Morgan Stanley’s SaaS index falling 15% year to date, as investors reprice companies exposed to AI-enabled workflow automation.
- Anthropic gains a clearer route to extend Claude Code-like capabilities to nontechnical users, though prompt-injection vulnerabilities remain a practical constraint on enterprise deployment.
Second-order effects
- SaaS vendors will face greater pressure to demonstrate that their products offer durable workflow, data, and customer-integration advantages beyond tasks a general-purpose AI assistant can perform.
- The market’s focus may shift from broad AI product announcements to evidence of adoption and defensibility, especially after Claude Cowork later added vertical plugins and customization tools the expansion into specialized workplace plugins.
Third-order effects
- If these reassessments persist, public-software valuations may increasingly separate companies that own embedded business workflows from those whose core features appear readily reproducible by frontier-model assistants.
- The episode points to a more contested application layer: AI labs can move from supplying models to competing for end-user work, while security and reliability limits determine how quickly that shift reaches enterprise use.
The trend: Frontier AI providers are moving from developer tools into workplace software, forcing SaaS investors to re-evaluate which application features remain defensible.