Key Takeaways:
Goldman Sachs strategist Ben Snider warned April 13 that AI disruption fears could weigh on growth stocks for years. Servicenow fell 48% and Salesforce dropped 36% YTD as per-seat licensing models face AI-driven “seat compression,” according to Yahoo Finance author Brian Sozzi’s reporting. Meta, Amazon, and Alphabet are positioned to recover first as Goldman targets selective exposure heading into 2027.
AI Fears Drive Software Stock Collapse in 2026, Goldman Sachs Strategist Warns No Quick Rebound
The warning, reported on by Yahoo Finance’s Brian Sozzi on Monday, lands as software equities are having a rough 2026. The report highlights how Servicenow is down 48% year-to-date. Salesforce has shed 36%. Docusign is off 42%. These declines are not random. The report explains that investors are pricing in “seat compression,” a scenario where a single AI agent replaces multiple human software users, gutting the per-seat licensing revenue that SaaS companies have built their business models around.
Sozzi details that the sector has lost roughly $2 trillion in market capitalization this year. Snider’s note, published by Goldman‘s U.S. Portfolio Strategy team, identified the core problem plainly: resolving investor uncertainty “will likely require evidence that AI is not displacing existing business models.” Until that evidence arrives through clean earnings beats and improving unit economics, share prices in vulnerable sectors are unlikely to find a floor.
In Sozzi’s report, Citi analyst Tyler Radke echoed Goldman’s concern, noting that worries about “software application architecture, business model durability and terminal value” could deepen in the coming months. Still, the Yahoo Finance editorial explains that private AI companies are projected to generate more than $100 billion in net-new revenue, pulling ahead of traditional application software on growth metrics.
The ‘SaaSpocalypse’ and What Goldman Is Watching
The Goldman note builds on the firm’s March 2026 report titled “Will AI Eat Software?” That 31-page analysis concluded AI is unlikely to fully displace software but will force major architectural change around large language models and autonomous agents. Incumbents hold some advantages through proprietary data and entrenched workflows, but the window to adapt is not open indefinitely.
Three large-cap names got a partial exemption in Snider’s framework. Meta Platforms, Amazon, and Alphabet are positioned to “regain their growth stock stride” on the back of strong expected results in 2026 and 2027. Their scale and AI integration give them a credible path that smaller SaaS platforms cannot yet claim.
The broader Magnificent Seven, however, is struggling, the Yahoo Finance report explains. JPMorgan strategist Mislav Matejka, quoted in Sozzi’s editorial, says the group is no longer performing its historical safe-haven role relative to the S&P 500. Only Amazon and Alphabet are marginally positive year-to-date. Tesla is down roughly 23%.
Capital is rotating toward sectors with physical assets, including data centers and infrastructure, where exposure to pure software disruption is lower, and AI infrastructure spending remains a direct tailwind.
Public Skepticism Adds Pressure Beyond Wall Street
Goldman’s institutional caution has a counterpart in public opinion. A Quinnipiac University poll surveyed 1,397 U.S. adults and found 80% are concerned about AI, with 70% believing it will reduce job opportunities. That figure is up sharply from 56% in Quinnipiac’s April 2025 poll.
Trust in AI-generated information remains thin. 76% of respondents said they trust AI outputs only “hardly ever” or “some of the time.” A separate NBC News poll found 57% of registered voters believe AI risks outweigh benefits.
Opposition to AI data centers is also hardening. 75% of Americans oppose having one built in their community, with 72% of opponents citing higher electricity costs and 64% pointing to water consumption. That local resistance is producing real project delays at a time when hyperscalers are still pushing capital expenditure projections higher for 2026.
74% of the poll’s respondents said the government is not doing enough to regulate AI, and 76% said businesses lack sufficient transparency about their AI use.
The tension the Quinnipiac data captures is real: personal AI tool usage is climbing, with 51% of respondents reporting they have used AI for research, up from 37% in 2025. But adoption is running well ahead of trust. That gap, combined with Goldman’s call for prolonged valuation pressure on growth stocks, suggests the AI cycle is entering a phase where skepticism, not enthusiasm, drives the narrative.
