A year ago, tech executives warned of mass AI-driven job destruction. Dario Amodei, CEO of Anthropic, said half of entry-level white-collar jobs could vanish. OpenAI’s Sam Altman foresaw the end of entire job categories. The Guardian’s analysis, published August 12, asks the question nobody expected to be asking: where’s the carnage?
🔍 THE BOTTOM LINE
The headline unemployment numbers haven’t moved much. But underneath them, AI is reshaping who gets hired, what skills matter, and how roles are structured. The jobs apocalypse may not have arrived, but the jobs transformation has — and it’s harder to measure.
What the Data Actually Shows
Research from the Stanford Institute for Economic Policy Research shows that since 2022 — the year ChatGPT launched — the unemployment rate for the 20 per cent of workers most exposed to AI rose by 0.77 percentage points. That’s actually less than the 0.85 percentage-point increase for the least-exposed workers. In other words, AI-exposed workers aren’t losing ground faster than anyone else.
“Employment trends in the occupations where we would expect to see the impacts first are largely stable,” said Erika McEntarfer, fellow at the Stanford Institute and co-author of the report. “It took decades for the computer revolution to fully transform labour markets in the workforce, and what we’re seeing right now looks a lot like that.”
AI could be a factor in recent graduates’ rising unemployment, which hit 5.6 per cent compared with the national average of 4.2 per cent earlier this year. But the Stanford report notes that remote work and the unwinding of pandemic-era overhiring probably also played roles. Disentangling AI’s effect from everything else is the core challenge.
The Challenger Context
The Challenger, Gray & Christmas July report adds nuance. US-based employers announced 33,429 job cuts in July — the lowest monthly total in two years. AI was cited as the reason for 10,970 of those cuts, or 33 per cent, making it the leading cause for the fifth consecutive month. Through July, AI has been cited in 112,713 job cut announcements, roughly 24 per cent of all cuts this year.
But here’s the counterpoint: total job cuts through July are down 41 per cent from the same period in 2025. Hiring has increased 25 per cent year-over-year. “While AI is shifting the labour market, it is not dismantling it,” said Andy Challenger, chief revenue officer for the firm.
The ambiguous nature of AI-attributed cuts was highlighted in July when Challenger noted that “naming AI in a layoff announcement can win over investors while pushing current and prospective employees away.” The messaging has swung from hedging to aggressively citing AI — which means the layoff numbers may overstate AI’s direct role.
The Rising Bar, Not the Shrinking Pool
The clearest signal in the data isn’t about job destruction. It’s about job transformation. About 74 per cent of employers consider AI skills a strong advantage or requirement, with 13 per cent requiring them company-wide, according to ZipRecruiter’s latest employer survey. Half of polled employers expect candidates to already be practical or advanced AI users on day one.
“The clearest trend line is a rising bar rather than a shrinking pool,” said Nicole Bachaud, a labour economist at ZipRecruiter. “The labour market challenge for workers is increasingly about skills-matching rather than pure job scarcity.”
Employers are simultaneously adding and cutting within the same functions — tech, customer support, business operations — which Bachaud says signals that “employers are still figuring out the exact skill set needed for success.”
This is consistent with what PwC’s AI Jobs Barometer has tracked: jobs requiring AI skills offer an average 62 per cent wage premium and are growing nearly eight times faster than the overall labour market. The most AI-exposed companies have tripled their lead in workforce productivity growth since 2022. Far from being a job killer, PwC suggests AI may be a job expander when used to unlock growth.
The Career Ladder Is Compressing
One of the most striking findings from PwC: AI-exposed junior roles are seven times more likely to demand traditionally senior skills like leadership and strategic thinking, compared to the least-exposed junior roles. The traditional career ladder is compressing. “Seniorised” entry-level roles — those that require judgment and decision-making above what was historically expected — have grown 35 per cent since 2019, even as overall early-career postings have flatlined in highly AI-exposed sectors.
This means the jobs aren’t disappearing. They’re getting harder. And the skills required to get in the door are shifting from technical execution toward the things AI can’t do well: empathy, judgment, creative problem-solving, and the ability to evaluate AI output critically.
As we’ve covered in our reporting on the entry-level career pipeline, the concern isn’t that jobs vanish overnight. It’s that the first rung on the ladder gets higher while the training ground that used to prepare people for it gets smaller.
The Freelance Future?
Paul Osterman, professor emeritus at MIT and author of the newly released book Disposable Workers, told the Guardian that more employers will likely turn to contractors and freelancers as they figure out the right mix of skills for an AI-shaped future. About 35 per cent of the US workforce is already considered “easily replaceable,” according to his research, and AI could exacerbate that by making it easier to swap workers in and out of project-based roles.
This matters for New Zealand, where the gig economy is already a significant portion of the labour market. If AI accelerates the shift from permanent employment to contract work, the implications for income security, career development, and collective bargaining are substantial. The IDC/Deel report found that 76 per cent of NZ organisations said fewer on-the-job development opportunities now exist for junior employees — a figure higher than any other country surveyed.
❓ FAQ
Has AI actually caused mass unemployment? No. The Stanford research shows AI-exposed workers’ unemployment has risen slightly, but less than workers not exposed to AI. The mass displacement predicted by some tech executives has not materialised in the aggregate data.
Is AI still affecting jobs? Yes, but in subtler ways. It’s changing what skills employers demand, compressing the career ladder, raising the bar for entry-level roles, and potentially shifting more work toward contract and freelance arrangements.
What should workers do? Develop AI fluency. The ZipRecruiter data shows 74 per cent of employers see AI skills as an advantage or requirement. PwC finds a 62 per cent wage premium for AI-skilled roles. The advice from economists is consistent: learn to work with AI, not against it.
Does this mean the fears were overblown? It’s too early to say. The Stanford researchers compare AI’s trajectory to the computer revolution, which took decades to fully transform labour markets. The current data shows stability, not stasis.
🔍 THE BOTTOM LINE
The AI jobs apocalypse didn’t show up on schedule. What arrived instead is something messier and more interesting: a labour market where the jobs still exist, but the rules for getting them have changed. The bar is higher, the skills are different, and the old career ladder is being rebuilt in real time. For workers, the message from the data is clear — adaptability, not panic, is the right response.
📰 Sources
- The Guardian — AI was supposed to destroy jobs. Where’s the carnage?
- Stanford Institute for Economic Policy Research — What is really happening to jobs?
- Challenger, Gray & Christmas — July 2026 report
- PwC — AI Jobs Barometer
- ZipRecruiter — Employer survey data
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
