Companies Are Cutting Jobs Over AI They Haven't Even Deployed Yet

Companies are cutting jobs over AI they haven’t deployed yet, so a job search strategy for 2026 should treat the hiring slowdown as a product of executive caution rather than AI replacing workers. In 2025, companies directly pointed to AI in announcing 55,000 job cuts, according to Challenger, Gray & Christmas as reported by CBS News, and a Harvard Business Review analysis argues that these layoffs are driven by AI’s potential rather than its performance.

Most of those companies haven’t actually deployed AI to replace the work those employees were doing. They cut headcount based on what AI might do, not what it’s doing now. Executives who read the same headlines about AI you did are preemptively trimming teams, and the jobs are disappearing before the technology arrives. This is the single most important thing to understand about the job market right now.

What is the “AI washing” problem in layoff announcements?

AI washing is blaming layoffs on AI that hasn’t been deployed. The pattern is consistent. A company announces layoffs. The press release or internal memo mentions AI as a driver. Analysts and media run with it. But when you look at what actually changed operationally, the AI implementation is often in pilot mode, planned for next year, or vaguely described as “efficiency gains from automation.”

The HBR piece draws on a December 2025 survey of 1,006 global executives. As summarized by Above the Law, about 60% of them had reduced headcount in anticipation of AI efficiency gains, while just 2% had made large reductions based on actual AI results. The disconnect exists partly because “AI” has become a socially acceptable reason to cut costs. It sounds forward-thinking. It implies strategic intent rather than financial pressure.

Other evidence points the same way. A Fortune roundup cites the Remote Labor Index, which found that less than 4.5% of jobs that could be done entirely over the internet could be adequately completed by AI agents, and notes that Klarna’s much-publicized AI customer service claims were followed by renewed hiring.

A separate MIT study, reported by Fortune in April 2026, scored AI output on a 1 to 9 scale where a 7 meant “minimally sufficient,” useful as is with no edits. As of late 2025, models reached that bar on roughly 65% of tasks, and superior output (a 9) never exceeded 50% success. In other words, the technology is passable on most tasks and strong on far fewer, which is a thin basis for replacing whole roles.

Amazon CEO Andy Jassy said in a 2025 memo that he expects generative AI and agents to reduce the company’s total corporate workforce over the next few years. Pinterest announced it would cut 15% of its workforce in January, with a spokesperson citing its “AI-forward strategy.”

What is actually happening in the job market?

The numbers tell a story of caution, not crisis.

Indeed Hiring Lab’s March 2026 report describes the “low-hire/low-fire” dynamic as entrenched. Net job creation has been minimal for over a year. Companies are holding onto the employees they have but not adding new ones. The economy added 178,000 jobs in March, but the three-month average is around 68,000 per month, and February was a net loss of 133,000.

The February 2026 JOLTS data showed the same pattern: a hires rate of 3.1%, the lowest since April 2020, and a quits rate of 1.9%. Workers are afraid to leave, and employers are in no rush to fill open positions.

For young job seekers in AI-exposed fields, there are early signs of an effect. An Anthropic study by economists Maxim Massenkoff and Peter McCrory, as summarized by People Matters, found that the monthly rate at which workers aged 22 to 25 move into high-AI-exposure occupations is about 14% below its 2022 level. The authors call the result only barely statistically significant, and they find no broad rise in unemployment in exposed occupations. The pattern looks like slower hiring rather than job losses. These aren’t jobs that AI is clearly doing right now. They’re jobs that companies may believe AI will eventually handle, so they may not be backfilling when people leave.

The practical result: fewer postings, more applicants per posting, longer searches, and more ghosting. The March 2026 BLS data showed long-term unemployment (27+ weeks) hitting 1.8 million people, up 322,000 from a year ago. A quarter of all unemployed Americans have been looking for over six months.

Why does your job search strategy in 2026 need to account for this?

Because the market is contracting on AI anxiety, so retraining and waiting won’t help much, while direct outreach does. If the market were contracting because of actual AI deployment, the playbook would be different. You’d retrain. You’d pivot to roles AI can’t do. You’d wait for the dust to settle.

But the evidence suggests the market is contracting largely because of AI anxiety, which is a different problem. Companies are cutting in anticipation, not in response to results. They may be pulling back hiring across the board, including in roles that have little to do with AI, because the general mood is caution. When the C-suite says “we’re investing in AI,” that can ripple through the org chart as fewer approved headcount requests, longer hiring timelines, and more positions frozen.

Your job search strategy in 2026 needs to account for this specifically. Traditional application channels are more clogged than ever. Ashby’s May 2026 data says applications per hire have tripled since 2021, to more than 300 per hire, and Greenhouse’s CEO told Fortune that 10 roles drew 2,539 applicants. Submitting an application and waiting for a response is asking a system that’s already slow and overloaded to notice you among hundreds of other candidates.

Direct outreach changes the equation. When you contact a hiring manager directly, you bypass the clogged pipeline. You’re not competing in a pile of hundreds of applications. You’re having a conversation with the person who has the authority and the motivation to fill the role. The hiring manager knows whether the position is real, whether budget has been approved, and whether they’re actually looking, all things that a job board listing can’t tell you.

How does direct outreach work in a cautious market?

In a market shaped by AI-driven caution, the hiring managers who are actually filling roles tend to be specific about what they need. They’ve fought for the headcount. They’ve justified it to leadership. They’re motivated to find someone quickly before the budget gets frozen.

That’s why direct outreach is worth the effort in this particular market. When a manager has an approved, funded position, they want to fill it fast. An unsolicited message from a qualified candidate who clearly understands the role saves them time. It gives them a candidate outside the ATS pile who they can evaluate directly.

The data on this points one way: Gem’s hiring data, drawn from 165 million applications and 1.2 million hires, shows that sourced candidates are eight times as likely to be hired as inbound applicants, and that direct sourcing accounts for about 11% of hires from 2.6% of applications. That is recruiter-led sourcing rather than job-seeker outreach, but it shows that being reached out to or introduced beats sitting in the pile. In a tighter market, there are fewer openings competing for attention and more motivation to fill the ones that exist.

Networking for job seekers in this environment means something more targeted than “attend events and hand out business cards.” It means identifying the 20-30 companies most likely to need your skills, finding the specific managers who’d be hiring, and reaching out with something relevant to say. A message that references a recent product launch, a team restructuring, or a specific challenge the company is solving beats any template. The research takes time. The payoff makes it worth doing.

What should you not do in a market shaped by AI anxiety?

Two common reactions to the AI-anxiety market are both counterproductive.

The first is panic-applying. If companies are hiring less, the instinct is to apply more. But volume-based strategies were already broken before the market tightened. AI auto-apply tools have made the application pile noisier than ever. Adding more noise doesn’t help.

The second is waiting. “I’ll wait until the market improves” sounds reasonable until you realize that the companies creating the slowdown don’t have a timeline for when AI will stop being their excuse. This holding pattern could last years. Indeed Hiring Lab’s 2026 forecast put end-of-year unemployment in the 4.1-4.8% range and projected job openings between 6.8 and 7.4 million, numbers that suggest a functional but cautious market, not one about to snap back to boom-era hiring. Sitting out means missing the openings that do exist right now, and those openings still favor candidates who show up with a direct message rather than a portal submission.

How should you adjust your job search strategy for 2026?

The market is being reshaped by fear of AI more than by AI itself. That fear is real, and it’s affecting hiring decisions across industries. But it doesn’t change the fundamental math of how people actually get hired.

Your job search strategy in 2026 should start with this reality: job boards and mass applications were already inefficient, and the AI-anxiety slowdown has made them worse. Companies are processing applications more slowly. Meanwhile, direct outreach, warm introductions, and hiring manager contact are worth more than ever because the managers who do have approved headcount are motivated to fill fast before the budget shifts again. The window for any given opening is likely shorter, which rewards the candidates who show up in someone’s inbox rather than waiting in line at the ATS.

The hard part is the research: figuring out who’s actually hiring, who the decision maker is, and what to say that shows you’ve done your homework. Angld.ai automates that research-to-outreach pipeline so you can focus on conversations with people who have real, funded positions to fill instead of sending applications into a market that’s running on AI anxiety.

Related: See how AI produces “good enough” work and what that means for your applications and learn why ghost jobs waste your time.