Until very recently, the narrative around AI was that the $600 billion of annual corporate capital expenditure (“capex”) fueling it was good for stocks in the short term. The companies receiving that money as new revenue (AI model makers, data center constructors, and the energy companies supplying them) would be the immediate beneficiaries. The efficiencies delivered by AI would be good for tech and non-tech companies alike. The Big Tech hyperscalers have always argued that the demand from their revenue-paying clients far exceeded their ability to supply AI services.
That narrative was turned on its head in the past 24 hours as it dawned on traders that AI also has the ability to reduce the revenues of a vast range of adjacent tech companies.
The argument—advanced by Palantir CEO Alex Karp and CTO Shyam Sankar on their recent earnings call—is that AI is now so good at writing or managing enterprise software that it threatens to make irrelevant a range of tech companies that have, for years, enjoyed recurring revenues by providing enterprise apps to companies on a software-as-a-service (SaaS) basis.
That led to a widespread selloff of tech stocks, wiping away $300 billion in market cap in a single session.
S&P 500 futures were flat this morning after closing down 0.84% last night.
SaaS companies took major hits: Microsoft closed down 2.87%; SAP was down 3.29% this morning on the German market; Salesforce lost 6.85% yesterday and was further down in overnight trading; ServiceNow was down 6.97% yesterday and was marginally lower overnight also.
Palantir’s Sankar said on the call that his company’s “AI forward deployed engineer” product—which allows clients to manage software and codebases through natural language commands—is able to reduce the time it takes to complete “complex SAP ERP migrations” from “years of work” to “as little as two weeks.” (ERP stands for “enterprise resource planning,” and it refers to a service offered by SAP around helping companies transition from aging legacy systems into new ones.)
Karp added: “In the American market, we have inbound [requests from clients] where people have already seen proof points at other companies and not on one use case. [There is] a myriad of use cases.”
Jefferies analysts Akshat Agarwal and Ayush Bansal—who focus on Indian enterprise tech companies—published a note this morning arguing that AI has the ability to reduce the revenues of a wide range of tech companies:
“Anthropic’s Cowork plug‑ins and Palantir’s claims of faster SAP migrations highlight how AI could potentially erode application service revenues for IT firms. With application services being 40% to 70% of revenues [for tech companies in India], IT firms face growth pressures. Consensus growth estimates don’t fully reflect this, posing downside risk to valuations,” they warned. Claude Cowork is like a general purpose work assistant that can organize tasks and files autonomously.
“Software stocks have been correcting on the back of this, however, we believe the impact of this could extend well beyond software—potentially disrupting downstream application‑managed services (AMS) revenues for IT services firms.
“Our checks suggest that use of AI is clearly compressing migration timelines which in turn may drag application implementation revenues for IT services firms,” they added. “AI is going to be drag on revenue growth of IT firms over the next one to two years.”
Ed Yardeni of Yardeni Research made a similar argument in a note to clients: “Software stocks were especially hard-hit because Anthropic rolled out new tools for its Cowork product. It’s too soon to tell how useful the new tools will be, but investors decided to cut the valuation multiples of software stocks.”
SAP refuted Karp’s point of view. “AI agents will massively push the boundaries of the performance of SaaS solutions, but not replace them,” the company told Fortune. A spokesperson argued that SaaS vendors will be the primary beneficiaries of AI because SaaS vendors have the data, the processes, and semantics that AI agents need to work.
“Companies still need a trusted source of truth to run the business. AI agents need clean, structured data and proven processes to produce reliable results,” the spokesperson said.
Here’s a snapshot of the markets ahead of the opening bell in New York this morning:
- S&P 500 futures were flat this morning. The last session closed down 0.84%.
- The STOXX Europe 600 was down 0.18% in early trading.
- The U.K.’s FTSE 100 was up 0.83% in early trading.
- Japan’s Nikkei 225 was down 0.78%
- China’s CSI 300 was up 0.83%.
- The South Korea KOSPI was up 1.57%.
- India’s Nifty 50 was flat.
- Bitcoin declined to $76K.
