AI Drives Tech Layoffs as Job Data Lags Reality

Tech companies including Meta, Amazon, and Oracle have announced thousands of job cuts in 2025 and 2026, explicitly citing AI-driven efficiency gains — even as National Economic Council Director Kevin Hassett told CNBC on May 11, 2026 that there is “no sign in the data” that AI is costing workers their jobs. The disconnect between official economic readings and corporate announcements has sharpened into one of the defining tensions of the current AI deployment cycle.

What Companies Are Actually Saying

The gap between government data and corporate action is stark. Meta plans to cut approximately 8,000 employees — roughly 10 percent of its workforce — with a human resources leader describing the move as necessary to “run the company more efficiently” and “offset the other investments” it’s making, according to Wired. Those cuts come on top of roughly 25,000 positions Meta has eliminated over the past four years.

Block went further. The payments company slashed its workforce by nearly half in February 2026, CNBC reported, explicitly citing a pivot to smaller teams using AI to handle work previously requiring larger headcounts. Amazon and Oracle have made similar announcements in the same period.

Coinbase CEO Brian Armstrong offered a candid illustration of the shift. In a May 5 post, Armstrong described being able to “ship in days what used to take a team weeks,” adding that non-technical teams are now shipping production code and that “many of our workflows are being automated.”

The pattern is consistent: companies frame reductions not as responses to poor performance, but as deliberate restructuring enabled by AI tools.

Inside Meta: Morale at a Record Low

The human cost of these announcements is playing out visibly inside Meta. Wired spoke with 16 current and former employees across a range of roles, all of whom declined to be named due to company policies barring unsanctioned press contact.

“Everyone is unhappy; the only people who are not unhappy are, literally, executives,” said one employee working on Instagram. A policy staffer described the atmosphere as “a bit ‘over it’ — lack of connection to the mission, upcoming layoffs, American employees being used to train the AI models that will replace them.”

Meta has also installed corporate software on employee computers to monitor activity for AI training purposes, according to those same employees — a detail that has compounded resentment. In the UK, some workers have begun organizing to form a labor union in response.

The financial picture makes the morale collapse more striking. Meta is reporting record-high profits even as it reduces headcount, a combination that has made the layoffs feel less like necessity and more like a choice to employees who remain.

Hollywood Writers Pivot to AI Training Work

The displacement is not limited to tech. A Hollywood writer and showrunner — who creates prime-time TV for Paramount, Hulu, and the BBC — described in a first-person account for Wired how the entertainment industry’s post-strike stagnation pushed them into AI training work to cover lost income.

The writer works under anonymous platform IDs like “ri611” for companies including Mercor, Outlier, Taskify, Turing, and Micro1. Tasks range from evaluating chatbot tone and annotating video timestamps to red-teaming large language models for safety vulnerabilities.

The 2023 Writers Guild strike was partly fought to prevent studios from replacing writers with AI. When it ended after nearly five months, the writer says, “the entertainment-industry carousel never gained back its momentum.” By early 2025, with a producer defaulting on a six-figure payment owed for a created TV show, they began looking for alternative income — and found it in the same AI systems the strike was meant to constrain.

The irony is not lost on the writer. Former TV professionals are now, effectively, providing the labeled data and safety evaluations that make AI writing tools more capable.

Benefits Cuts Follow Headcount Reductions

For workers who keep their jobs, the financial squeeze is extending into benefits. Wired documented three recent cases where companies reduced non-wage compensation, with AI investment cited as a contributing rationale.

• TTEC, a Texas tech consulting firm, suspended its discretionary 401(k) match for 16,000 employees through at least the end of 2026. An internal memo viewed by Business Insider indicated the company plans to redirect funds toward AI certifications, tools, training, and automation.
• Deloitte is reportedly cutting paid time off, halving parental leave, and eliminating a $50,000 reimbursement for family planning services including adoption, surrogacy, and IVF — but only for workers in internal roles such as admin, IT support, and finance, while leaving client-facing employee benefits intact.
• Zoom reduced parental leave for birthing parents from 22 weeks to 18 weeks.

Joan C. Williams, a professor at UC Law San Francisco and author of several books on work culture, told Wired that Deloitte’s approach “treats people differently based on the type of job they’re in, and cutting any mother down to eight weeks of paid leave is just outlandish.” She added: “When labor is tight, employers are more generous. But once the power shifts, the benefits contract.”

What This Means

The Hassett statement — that there is “no sign in the data” of AI-driven job losses — reflects a real limitation of aggregate employment statistics. Macro data captures net employment levels; it does not capture the composition of who is being hired, what skills are valued, or the quality of positions being created versus eliminated. A company can cut 8,000 mid-level roles and hire 500 AI engineers, and the headline unemployment figure may barely move.

What the corporate announcements reveal is a deliberate restructuring of labor costs, not a cyclical downturn. Meta, Block, TTEC, and others are explicitly connecting workforce reductions to AI efficiency gains in their own communications — not to revenue shortfalls or market contractions. That framing matters: it signals that these cuts are structural, not temporary.

The benefits erosion adds a second layer. Even for workers not facing layoffs, companies appear to be recalibrating what employment includes — reducing parental leave, retirement contributions, and family planning support while simultaneously investing in AI tooling. Williams’ observation that benefits contract when employer power increases is a useful frame: AI-driven productivity gains shift negotiating leverage toward employers at precisely the moment workers have fewer alternatives.

The Hollywood writer’s account illustrates where some of that displaced labor ends up: in low-wage, anonymized gig work that directly feeds the AI systems that displaced them. That loop — skilled workers training the models that reduce demand for their skills — is likely to intensify before any policy or labor response catches up.

FAQ

Why does government data show no AI job losses if companies are announcing layoffs?

Aggregate employment statistics measure net job levels across the full economy, not the composition or quality of roles. A wave of mid-level cuts at large tech firms can be offset by hiring elsewhere, leaving headline unemployment figures stable even as specific occupational categories shrink significantly.

Which companies have announced AI-related layoffs in 2025 and 2026?

Documented cases include Meta (approximately 8,000 cuts, or 10 percent of its workforce), Block (nearly half its workforce, February 2026), Amazon, and Oracle. TTEC suspended 401(k) contributions for 16,000 employees while redirecting funds to AI tools and training.

What jobs are displaced Hollywood and creative workers doing instead?

According to a first-person account in Wired, former TV writers are working as AI trainers for companies like Mercor, Outlier, Taskify, Turing, and Micro1 — performing tasks such as evaluating chatbot responses, annotating video data, and red-teaming language models for safety vulnerabilities.