Traditional CRM helped define the modern enterprise stack. Now, investors are starting to wonder whether AI-native tools will unbundle it—and they’re pricing that fear in fast.
In the latest bout of anxiety, software stocks sold off sharply, with Reuters estimating the S&P 500 Software & Services index has shed about $1 trillion in market value since Jan. 28, 2026, as traders recalibrated what “defensible” looks like in an era of copilots and agents.
A big part of the narrative is that AI-native challengers aren’t just adding features to CRM—they’re trying to replace the workflow entirely. One of the names pulling attention is Day AI, a startup founded by former HubSpot executives Christopher O’Donnell and Michael Pici. Day AI is pitching a system that can absorb company context quickly and behave more like an always-on “chief of staff” than a traditional database, helping with meeting prep, pipeline updates, and internal Q&A—at least, that’s the promise.
Why the selloff hit CRM especially hard
Markets tend to punish incumbents when a new interface becomes the main entry point. In CRM, that interface could be an agent that can answer “What’s the deal risk?” or “Draft a follow-up based on the last call,” without anyone opening a dashboard.
That’s the fear showing up in prices. HubSpot shares are down roughly 71% year over year, while Salesforce is down about a third over the same period. Across the sector, several customer-engagement and workflow names have also fallen sharply—fueling the argument that Wall Street is treating older software models like potential value traps.
Wedbush analyst Dan Ives told Bloomberg the slump is “unlike anything I’ve ever seen,” as investors begin valuing companies as if AI has already done the damage—even though the industry is still early in figuring out what “agent-first” software really means.
The Anthropic spark: agents, plug-ins, and category panic
The anxiety flared again after Anthropic rolled out plug-ins for its Claude “Coworker” agent, with examples that automate legal work like contract review and compliance. For markets, it was less about one product announcement and more about what it signaled: if AI agents can learn tools, act on your behalf, and operate across systems, then entire software categories could be at risk—not just individual features.
That’s the backdrop for the “CRM extinction” talk. The bolder version of the story says legacy platforms won’t survive. The more realistic version says they’ll survive, but they’ll be forced to change shape: fewer forms, fewer tabs, fewer dashboards—more agent experiences wrapped around data, permissions, and workflows.
The incumbent response: “We’re building agents too”
Salesforce is pushing exactly that counterargument. The company points to Agentforce, saying it has logged 18,500+ deals since launch, including 9,500+ paid deals, and that Agentforce ARR is up 330% year over year—numbers Salesforce describes as its fastest organic ramp.
The pitch from incumbents is straightforward: AI without deep data context, governance, and security turns into guesswork—or worse, spammy “spray and pray” automation. In that framing, the winners won’t be the flashiest chatbot. They’ll be whoever can combine trusted customer data + workflow permissions + action-taking agents without breaking compliance or trust.
So… is this really the “death of CRM”?
Probably not. But the pressure is real.
AI-native startups are attacking CRM where it’s weakest: manual updates, stale fields, and the endless admin work that makes sales teams feel like they’re feeding the system instead of the system helping them. The best AI challengers also have an advantage that’s hard to ignore—many are built by people who used the legacy tools, know the pain points, and can redesign the workflow from scratch.
What investors are reacting to isn’t that CRM disappears overnight. It’s that the value moves up the stack—from “a place to store customer records” to “a system that understands customer context and completes work.” If that shift happens, today’s giants may still be around, but the CRM people recognize could look very different.