Fear Of Missing Out or FOMO a term which demonstrates as pressure to move quickly, and keep pace with peers/competitors. The year of 2026 has become the year where the AI industry's biggest claims are being tested against reality, and with no surprise the results are mixed. The industries’ revenues are climbing to genuinely unprecedented levels. At the same time, a growing number of evidences show that companies are walking back on some of AI's most aggressive promises, the most recent example of 1 such company is Ford.
The growth case
Forbes' AI 50 list mentions that companies on this year's list have raised around a combined of $305.6 billion, with OpenAI and Anthropic alone accounting for 79% of the pool. Goldman Sachs Asset Management found that enterprises are rapidly deploying AI after a slow start, with the top 5% of companies now consuming three times more the tokens of the median company creating a gap which is only widening with time and not narrowing.
The adoption gap
But the AI use adoption breadth has not translated equally into value. McKinsey's global survey in 2025 found that 88% of organisations use AI in at least one business function, though many have not integrated it enough to capture enterprise level returns. A Forbes Tech Council article attributes this to "AI FOMO" adoption driven by competitive anxiety rather than a defined problem to solve.
See here It cites four recurring constraints slowing AI Execution: strategy drift, non calibrated risk appetite, Lack of specific policies regarding AI, & Governance.
Investors are recalibrating too
ING's notes that while AI remains a positive long-term story, heavy infrastructure spending is raising depreciation costs and cutting into share buybacks pointing to slower EPS growth and lower valuation multiples ahead, with forward Nasdaq valuations already near the low end of their historical range. ING flags Oracle, Nvidia, OpenAI, and Anthropic as carrying particular company-specific risk, and projects Oracle's EBITDA less capex could turn negative despite its scale. CBS News reported that tech selloff reflects "gnawing anxiety" over whether the trillions being spentGoldman Sachs estimates $7.6 trillion through 2031 on data centres alone will generate matching revenue. Public sentiment is part of that anxiety: Pew Research found 40% of American adults believe AI will be a negative societal force over the next two decades, versus only 16% who see it as positive.
The ex's promise: "AI will replace engineers"
This is where the gap between claims and results has been contrasting and most costly. Since 2023 & beyond, the narrative was that AI would heavily reduce and sometimes remove software engineering and other skilled roles, completely. Companies got carried away & also started acting on it. Now since we have crossed the mid of 2026, and some (of many) documented results are below, the most recent being (till the time of publishing this article)Ford which had cut roughly 5,000 jobs while introducing AI into vehicle quality inspections and design reviews, but then the recalls kept climbing when company issued 152 recalls last year, the highest of any U.S. automaker. Ford has since rehired 350 veteran engineers, including retirees, who now lead design reviews, train new hires, and work on improving the AI tools themselves. Ford's VP of vehicle hardware engineering, Charles Poon, said that the company had wrongly assumed that feeding AI its design requirements would produce a high-quality product without experienced human oversight, and CEO Jim Farley said the reversal has generated hundreds of millions of dollars in savings from declining warranty and recall costs. Other famous examples being IBM, Commonwealth Bank of Australia, Klarna etc.
The pattern isn't isolated. A Robert Half survey found roughly 29–32% of hiring managers who eliminated a role citing AI later rehired for that role or a comparable one, and 35.6% rehired more than half of those they'd let go with a third of those employers spending more on restaffing than they'd saved from the original layoffs. Forrester’s Research 2026 "Future of Work" report estimated 55% of employers regretted AI-driven layoffs, and Gartner projects half of all companies that cut jobs for AI reasons will rehire for similar roles within a year.
The bottom line
Usage, revenue, and enterprise compute demand are genuinely climbing, and serious capital continues to flow into the sector on the strength of real productivity gains. But the evidence is equally clear that AI has not yet delivered on some of its most aggressive promises like replacing skilled engineering judgment being the clearest example and that a share of 2024–2025 layoffs were premature bets that companies are now paying to unwind.
Way Forward
Now stating these facts let’s understand what’s the way forward :-
1. Treat AI adoption as an operating-model decision, not a headcount decision. The clearest thread across Ford, IBM, and Klarna is that each treated AI as a substitute for judgment rather than a tool that still needed judgment applied to it. Before cutting a role, the more durable question is the one the Forbes FOMO piece raises: what specific problem is this solving, who owns the risk if it's wrong, and where is human review non-negotiable? Ford's mistake, in its own VP's words, was assuming that feeding AI the requirements would produce a good outcome without oversight.
2. Don't cut the people who could catch AI's mistakes. IBM's AskHR case is instructive precisely because the failure rate was low (6%) — but that 6% included the judgment calls that mattered most.
3. Protect the entry-level jobs even while automating routine work. The Stanford HAI data on falling employment for developers at joining level is genuine.Junior roles do require a lot of routine work does get displaced, by AI. But IBM's own logic applies broadly: if you stop training juniors, there's no pipeline of future senior engineers who can supervise AI in five years. So a continued junior hiring, even at reduced pace, is very much a necessity.
4. Measure value created by AI, and not just adoption. McKinsey's 88% adoption of AI but lower integration gap suggests companies are optimising for "are we using AI" rather than "is this specific use case working."
5. Where trust and relationships are the product, use AI to support people, not replace them. In domains where output depends largely on human trust like sales, high-touch customer service, donor relations, clinical or ethical judgment the companies faring better IKEA's model of upskilling staff into AI-assisted consultants is a better example to follow.
None of this argues for pulling back from AI. It argues for the more boring, less exciting version of adoption shows with defined metrics, keep the humans who know what AI misses, and treat "AI replaces the role entirely" as a hypothesis to test on a small scale before it's a company-wide bet.
With this signing off
— Sudarshan Mishra
Founder, Easy Web Presence