Some of the largest and most profitable companies are cracking under the weight of staying in the artificial intelligence (AI) race.
In our prior reports (November 2025 and February 2026), we warned that the “losers” in the AI race (those without the free cash flow to sustain spending) would be forced either (1) to cut spending or (2) dilute investors to stay in the AI race.
However, recent corporate actions by Amazon (AMZN), Meta (META), and Oracle (ORCL), introduce a third option: layoffs. There have been many headlines about AI replacing humans for many jobs, and now we’re seeing it happen in real time.
The “AI Purge”
Instead of further diluting exasperated investors, the “losers” in the AI race are clearing the decks of everything not AI and laying off thousands of employees.
- “Amazon laying off about 14,000 corporate workers as it invests more in AI” – Oct 2025
- “Amazon laying off about 16,000 corporate workers in latest anti-bureaucracy push” – Jan 2026
- “Oracle Plans Thousands of Job Cuts in Face of AI Cash Crunch” – Mar 2026
- “Meta planning sweeping layoffs as AI costs mount” – Mar 2026
As Reuters notes, “Meta seeks to offset costly artificial intelligence infrastructure bets…” Similarly, Bloomberg notes Oracle plans “to ax thousands of jobs, among its moves to handle a cash crunch from a massive AI data center expansion effort.”
Just a Band-Aid For a Gaping Cash Flow Hole
Cutting employees to invest in chips and buildings might protect margins in the short-term but is a small band-aid for multi-billion-dollar cash flow hemorrhaging.
Meta plans to spend between $115-135 billion in capex in 2026, up from $72 billion in 2025. Meanwhile, analysts at JPMorgan and Bank of America estimate the reported layoffs could save the company $6-8 billion annually, or less than 6% of the midpoint of Meta’s capex estimate.
It is more of the same at Amazon. The 30,000 in combined layoffs are estimated to save the company $8 billion in 2026, or just 4% of its planned 2026 AI spend.
Oracle’s cuts look to be most impactful, which makes sense given that they have the least cash flow runway to remain in the AI race. The reported job cuts are estimated to save the company between $8-10 billion, which would represent 18% of its 2026 capex at the midpoint of estimates.
These cuts will no doubt free up some cash and allow each company to remain in the AI race a little longer, but job cuts provide only one-off cash flow savings. There is a limit to how many employees a company can lay off before seeing diminishing returns.
The Data Spend Is Even More Important
Ultimately, these companies will need to sustain high spending rates for much longer than they can afford to cut jobs. We think the AI race is still in its early stages.
Infrastructure is important, but spending on high quality-data, and not general internet data, will truly set AI systems apart from one another.
“These AI models are trained on publicly available data…they’re all trained on all of the data on the internet…But for these models to reach their peak value, you need train them not just on publicly available data, but you need to make private, privately owned data available to those models as well.”
– Larry Ellison during Oracle investor presentation
Every day, more and more people are realizing that the most essential ingredient in building reliable AI systems is high-quality data. The saying “garbage in, garbage out” applies to AI models as much or more than any other model. No matter how much infrastructure or large the language model, an AI system trained on unreliable data will be unreliable.
Until users can trust AI to reliably perform complex tasks, they will not likely be willing to pay much for access to AI. If users are not willing to pay substantially more for access to AI, the gigantic AI infrastructure spend will not earn its spenders an adequate return on investment. In which case, those stocks could plummet.
Consequently, we expect an entirely new AI spending spree to build or buy high-quality datasets that can truly endow machines with deep subject matter expertise. When AI can deliver correct, truthful answers to questions that require deep subject matter expertise, then it would be reasonable to expect an explosion in AI-driven revenue and profits. Until then, we see AI’s utility as far below what’s needed to generate an adequate return on the gigantic AI spend.
In other words, only the companies that can add the highest quality training datasets to industry-leading AI infrastructure will be positioned to win the AI race.
We believe that AI is forcing most big tech companies to incur Industrials-like levels of capex, which will lead to Industrials-like levels of return on invested capital (ROIC).
In that scenario, there’s no straight-faced argument for the AI tech giants to retain their current sky-high valuations.
This article was originally published on March 23, 2026.
Disclosure: David Trainer and Kyle Guske II receive no compensation to write about any specific stock, style, or theme.
Questions on this report or others? Join our online community and connect with us directly.