Tech Explained: Here’s a simplified explanation of the latest technology update around Tech Explained: Tech Jobs Are Vanishing. Is AI to Blame?? in Simple Termsand what it means for users..
In May 2025, Anthropic CEO Dario Amodei predicted that AI could eliminate nearly half of white-collar jobs and raise unemployment rates to 20 percent within the next five years. According to recent ADP data, the tech sector might already be taking that hit.
Information and business services shed a total of 62,000 jobs in January, marking four consecutive months of losses per the ADP National Employment Report.
But is AI actually responsible for these cuts? It may certainly be a contributing factor. Researchers at the Federal Reserve Bank of Saint Louis found that occupations with higher AI exposure experienced higher increases in unemployment rates between 2022 and 2025. And whether true or not, many tech CEOs cited AI as the reason for mass layoffs or hiring freezes last year.
The more nuanced question to explore might not be if AI is killing tech jobs but how. Conventional wisdom suggests that tech companies will hire fewer knowledge workers as AI makes them more efficient and even automates some workstreams entirely. We might see this at scale in the future, but there’s a more reasonable explanation for the current job market slowdown.
AI costs money. Lots of it. If you want to triple or quadruple your spending on it, the easiest place to find that money is payroll. The massive job losses might not be coming entirely from AI replacing human labor, but from companies who have gone all-in on AI and now need to right-size AI investments sitting on the P and L sheet.
How Is AI Impacting Tech Hiring?
AI is driving tech layoffs primarily through high operational costs rather than direct human replacement. Companies are cutting payroll to fund expensive AI line items like compute and integration. While AI increases productivity, the immediate job market hit comes from businesses cutting budgets to afford the massive investments required to stay competitive.
Why Have Tech Jobs Been Hit So Hard?
Some have pointed to tariffs and the overall weak economy to justify the slow hiring that plagued the end of 2025. Analysis from the Kansas City Fed does suggest that domestic companies have added fewer jobs in direct response to tariffs. Businesses also tend to pull back on hiring when they can’t affordably access debt. Between 2022 and early 2024, interest rates jumped by more than 5 percent. When the Fed cut rates for the third time in late 2025, it did so in response to the weak job market.
But if tariffs and recession-like economic conditions are to blame, other industries should be cratering at a similar rate. Yet ADP data suggests the slowdown is concentrated in the sectors most exposed to AI. Information and business services experienced the highest net job losses out of any service-providing industry, eliminating 41,000 and 107,000 positions respectively over the last 12 months.
There’s no doubt AI can make tech workers more productive. A board only needs a rough estimate of those potential gains to act on this belief. If corporate leaders think they can squeeze 20 percent more out of their workforce, they’ll pre-spend it by cutting budgets. The biggest budget they can cut is payroll.
The Economics of AI Are Hurting Tech Jobs
AI introduces a new line item that competes with people. That’s true even if the tools feel cheap at the seat level. If you keep the workforce flat and add AI, the real costs pile up in compute, usage and integration.
We’re only a few years deep into modern generative AI adoption. Many companies are still in the middle stages of implementing it, to say nothing of measuring its impact. Some might be shipping AI to keep up with market expectations as they scramble to figure out how it actually adds customer value. While software teams are experimenting with new workflows and wiring it into products, the AI bills — sky-high and unpredictable — keep flowing. Without disciplined management of your AI spending, the volatility of token-based costs can devour IT budgets quickly.
If they aren’t finding ROI from AI in new revenue, companies will look for it in costs. Return logically comes from payroll for two reasons. One, payroll is a heftier, faster-moving cost center more easily controlled than revenue. Two, AI is intended to eventually displace human talent. That creates pressure to cut headcount based on expected productivity improvement, even as employees are still seizing AI’s full potential.
Slow Hiring May Reverse as the AI Market Matures
The same force responsible for causing tech job losses could also eventually curb them.
AI vendors are in a race for market share with a new frontrunner emerging seemingly every week. We’ve yet to see what AI prices will look like once vendors stop burning cash and start acting like mature, profitable tech infrastructure providers. OpenAI, for example, is expected to burn $115 billion through 2029 as it funds the chips and data centers the company says it needs to sustain demand.
Typically, prices decline as an industry matures. I think the inverse will happen with AI. No vendor wants to be the first to raise prices because they’ll lose market share, but it will happen. AI workloads are eye-wateringly expensive to run. They guzzle energy and require advanced cooling solutions. Meeting worldwide demand for AI will require a $5.2 trillion investment in data centers by 2030, McKinsey estimates.
AI use cases will narrow as AI economics become more restrictive. Companies will start to determine which workloads are worth the cost to run, weeding out deployments that aren’t contributing to revenue. When companies pump the brakes on funding broad AI programs, payroll might finally stop getting squeezed. Finance can budget against specific initiatives with defined outcomes instead of vague or experimental investments that compete with headcount.
AI Will Cut Budgets Before It Cuts Your Job
With 90 percent of tech workers using AI at work, we likely haven’t reached the ceiling of its full productivity lift. We may see more churn before we see the revenue upside of AI-enhanced workforces.
In the meantime, how should mid- and early-career tech workers navigate this transition? I’ll offer the advice I’ve shared with my software developers in hopes that it applies more broadly.
First, don’t sit this one out. Assume everyone is using AI tools. The baseline has shifted. Make AI your pair programmer instead of your replacement.
Second, hone the skills AI doesn’t excel at yet. AI can generate code and even critique it, but it still struggles with debugging complex systems, reasoning about real-world constraints and exercising judgment in context. You’re the one who understands the system, the customer and the tradeoffs. You’re the one with veto rights over what gets shipped.
Third, if you love building software, don’t let the headlines scare you into another field. The industry still needs people who want to do the work. Developers who can show results from integrating AI into their workflows don’t look like expendable headcount.
And finally, although we’re not out of the woods yet, take assurance in knowing there’s a more practical explanation for the industry’s slow hiring and job losses: companies trying to justify their expensive AI bets to boards and investors. Not robots taking your seat.