🚨 AI Is Coming for Software—and One Top Fund Manager Says “Sell Before It’s Too Late”

In a market obsessed with buying the next artificial-intelligence winner, one veteran investor has taken the opposite bet—dumping much of the traditional software sector and warning that many of today’s biggest names may not survive the decade.

Nick Evans, who helps oversee roughly $12 billion at Polar Capital, says the industry that powered stock markets for years now faces an “existential threat” from the very technology investors once believed would strengthen it.

His fund’s performance—beating the vast majority of global tech peers over both one-year and five-year periods—has given weight to his stark message: the disruption from AI is only beginning.

This year has already drawn a sharp dividing line across the technology landscape.

That divergence reflects a deeper shift: investors are starting to believe AI may replace parts of the software business model instead of just enhancing it.

Tools such as those developed by Anthropic can now generate, adapt, and maintain code—tasks that once required expensive enterprise platforms and armies of developers.

Evans argues that application software—the everyday systems used for payroll, customer management, and document creation—is particularly exposed.

These tools have historically locked companies into long contracts. AI could break that model by allowing businesses to build customized solutions internally at a fraction of the cost.

Reflecting that view, Evans has exited positions in well-known enterprise providers such as SAP, ServiceNow, Adobe, and HubSpot.

“We won’t go back to these companies,” he said, signaling a decisive break from a sector long considered a cornerstone of growth portfolios.

His only meaningful exposure remains a small position tied to Microsoft, largely as a hedge through options rather than a full-scale endorsement.

Instead of software, Evans has leaned aggressively into the physical backbone of AI.

Seven of the fund’s ten largest holdings are semiconductor companies, led by Nvidia, whose processors have become synonymous with AI infrastructure.

He’s also bullish on:

In his view, the AI boom resembles a modern-day gold rush—where the winners are selling the tools, not the applications.

Beyond competition, Evans sees a financial squeeze forming inside software firms themselves.

Many employees are compensated with stock. If share prices weaken:

“We don’t believe current prices reflect the pressure on free cash flow,” he warned.

Evans isn’t entirely abandoning the sector. He has selectively added exposure to infrastructure software—the digital plumbing behind the internet—through firms like Cloudflare and Snowflake.

Recent earnings from companies such as Datadog and Fastly suggest demand for foundational systems is accelerating, even as traditional applications struggle.

Cybersecurity, he adds, appears relatively insulated—for now.

Still, these areas represent only a small slice of his portfolio.

Not everyone agrees with Evans’ bleak outlook.

Analysts at JPMorgan Chase recently argued that the sell-off may be overdone, predicting a rebound in select software names as markets digest the changes.

The debate underscores how uncertain—and emotional—the transition to AI-driven economics has become.

Evans draws a stark historical comparison.

Just as the internet devastated print media by removing barriers to publishing, AI may dismantle the traditional advantages of software vendors.

His conclusion: only a handful of companies will adapt quickly enough to remain dominant.

Investors, he says, must move fast.

The first phase of the AI boom rewarded companies that built software.
The next phase may reward those that enable intelligence to run everywhere—from chips to networks to energy grids.

For investors, the shift could redefine what “tech investing” means for years to come.