🔍 Read the full analysis: AI And New Graduate Jobs: Reading Beyond The Headlines on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Job-ad data from the UK and Australia show graduate and other hiring has weakened, with larger declines in occupations more exposed to AI. The patterns do not prove AI caused the falls: economic conditions and other labour-market changes are also plausible factors, and the scale and timing of AI’s employment effects remain uncertain.
Job adverts have fallen more sharply in occupations considered highly exposed to AI in recent UK and Australian data, raising concerns about graduate entry-level work. But the figures show a pattern, not proof that AI caused the decline: researchers cite other pressures on hiring, and the UK fiscal watchdog says the scale and timing of AI’s effects remain uncertain.
A Bank of England staff analysis by Haley Schlicht compared UK online job adverts with an AI-exposure measure for occupations. It found adverts down 15% in the most exposed third of occupations, compared with 6% in the least exposed third; the middle third fell 10%. Customer service adverts dropped 23% and administrative adverts 22%. The analysis used Office for National Statistics job-ad data and was published on the Bank Underground blog in August 2026.
Schlicht cautioned that the figures do not establish why hiring fell. Pandemic-era over-hiring, an economic slowdown, remote work, an energy shock and higher labour costs occurred around the same period. The analysis says “confident attribution remains premature”. Separately, the ONS estimated 691,000 UK vacancies in July to September 2026, down 21,000, or 2.9%, from the previous quarter. It said smaller firms may not be recruiting because of increased labour costs.
In Australia, graduate job postings fell almost 15% in 2025, with steeper declines in more AI-exposed roles, according to the source report’s account of Indeed data. The decline began in 2023, before widespread AI adoption, and apprenticeship postings—described as having low AI exposure—fell 46% from their peak. Those figures complicate any claim that AI alone explains weaker opportunities for new workers.
AI and the entry level · explainer
From career pyramid to diamond
Why the first rung of the ladder is where AI could show up first, and why the cause is not settled yet.
A workforce by level (schematic)
Pyramid: many juniors at the basedoing scheduling, correspondence and routine information processing UK job adverts: −15% vs −6%in the most vs the least AI-exposed third of occupations Diamond: hollow at the base“centred on experienced staff”
How the squeeze could work
- First jobsRoutine tasks: scheduling, correspondence, routine information processingthe base of the pyramid
- AI steps inSenior workers can do more of the work that once formed the bedrock of junior hiringBank Underground
- Fewer advertsUK online job adverts fell 15% in the most AI-exposed third, 6% in the leastBank Underground, ONS data
- DiamondThe pyramid may turn into a diamond, hollow at the baseBank Underground
- Not settledPandemic over-hiring, the business cycle, remote work and labour costs hit at the same time“confident attribution remains premature”
Source: Haley Schlicht, Bank of England, Bank Underground blog, Aug 6, 2026 (staff view, not necessarily the Bank’s), based on ONS online job adverts through early 2026; the pyramid-to-diamond idea is from research the post cites. The blocks are a schematic, not data. The question in the orange box is ours.
Graphic: Powered by Thorsten Meyer AI
The Shrinking First Rung
The pattern matters because customer service and administrative work can provide early-career experience, workplace training and a route into more skilled roles. If employers reduce junior hiring in these areas, graduates may have fewer opportunities to build practical skills and progress. The data cited here, however, do not show whether such changes are permanent or whether jobs are being eliminated, redesigned or shifting elsewhere.
Employers also have a longer-term interest in the pipeline. If fewer junior workers enter a profession, businesses could later face a shortage of staff with the experience needed for complex work. That is a risk raised by the source analysis, not a demonstrated outcome of the current job-ad figures. For graduates, the immediate issue is a weaker hiring market, especially in roles with routine information processing.
For universities and employers, the discussion is not simply whether students can use AI tools. The Conversation article by Armin D. Ashenberg argues that graduates need to judge what a good result looks like, identify where AI output falls short and improve it. That is an expert argument about skills, rather than evidence that employers have adopted a particular hiring standard.
Products Worth Considering

Dell XPS 13 13.4-Inch Touchscreen Laptop, Intel Core 5 320, 16GB RAM, 512GB SSD

Dell UltraSharp 27 Thunderbolt Hub Monitor (U2726DER), QHD

Dell XPS Tower 8960 Desktop, Intel Core i7-13700, NVIDIA GeForce RTX 4060, 16GB DDR5, 512GB SSD
As an affiliate, we earn on qualifying purchases.
Two Warnings, Different Evidence
The immediate discussion follows two published warnings. On November 15, Armin D. Ashenberg, a senior lecturer in law and emerging technologies at Western Sydney University, wrote in The Conversation about how AI could change graduate work. On November 13, The Telegraph reported that the head of the UK Office for Budget Responsibility had warned of “very painful” job losses from AI. The Telegraph article was behind a paywall, and the source report says it found no OBR or parliamentary record of the quoted remarks or another outlet directly reporting them. The warning should therefore be treated as The Telegraph’s report, not as an independently verified OBR statement.
The OBR’s published October 2026 Economic and fiscal outlook provides a separate, attributable account of its position. Its central forecast is for AI to add 0.2 percentage points a year to productivity growth by the medium term, while stating that the scale and timing of the effect are “highly uncertain.” The report cites ONS figures showing 30% of businesses reported using at least one AI technology in September 2026, up from 15% two years earlier.
The OBR also said it would monitor emerging evidence on productivity and employment. These points establish that official forecasts include a potential productivity gain while acknowledging substantial uncertainty; they do not confirm a specific forecast of AI-driven job losses.
““Confident attribution remains premature.””
— Haley Schlicht, Bank of England Data and Statistics Division
Cause and Scale Remain Unsettled
Job-ad declines do not identify their cause. The UK analysis compares changes across occupations grouped by AI exposure, but also lists economic and workplace shifts that could have reduced hiring. Australia’s graduate-ad decline began in 2023, before widespread AI adoption, while apprenticeship postings fell despite being described as barely exposed to AI. The available figures do not separate these effects or show how many roles were lost specifically because of AI.
The reported OBR warning also remains unverified in the material cited here. The source account says the relevant Telegraph story was paywalled and that no matching OBR speech, transcript or committee record was found. The OBR’s published forecast instead emphasizes uncertainty and says it will monitor the evidence. The data presented do not establish what individual employers plan to do, how graduate hiring will change over time, or whether reduced postings will translate into fewer hires.
More Hiring Evidence Needed
The next useful evidence will be updated job-ad and vacancy data, alongside information on actual hiring, redundancies and employer use of AI. The ONS vacancy series can show whether the UK market continues to weaken; it cannot by itself identify AI as the cause. Comparisons across industries and worker seniority, with consistent time periods and measures of AI exposure, would help clarify whether entry-level roles are changing differently from other jobs.
The OBR said it would closely monitor emerging evidence for its productivity and employment forecasts. Universities and employers will also face questions about how they prepare and train new workers, including whether graduates can evaluate and improve AI-generated work. For now, the data support concern about weaker entry-level hiring in some exposed occupations, but not a definitive claim that AI is responsible.
Key Questions
Are AI-exposed graduate jobs disappearing?
The cited data show job-ad declines in some AI-exposed occupations, not that those jobs are disappearing altogether. They also do not establish how many changes are due to AI.
What did the UK job-ad analysis find?
The Bank of England staff analysis found online adverts fell 15% in the most AI-exposed third of occupations and 6% in the least exposed third. Customer service adverts fell 23% and administrative adverts 22% over the period examined.
Did the OBR confirm the reported warning about painful job losses?
No such confirmation is established by the cited material. The warning was reported by The Telegraph; the source account says it found no corresponding OBR or parliamentary record. The OBR’s October forecast says the scale and timing of AI’s effects are highly uncertain.
Does the Australian data prove AI caused graduate postings to fall?
No. Graduate postings fell almost 15% in 2025 and fell more sharply in AI-exposed jobs, according to the source report’s account of Indeed data. But the decline began in 2023, before widespread AI adoption, and apprenticeship postings also fell substantially.
What evidence could clarify AI’s effect on entry-level work?
Updated vacancy and job-ad data, combined with records of actual hiring and employers’ AI use, could help. Researchers would need to distinguish AI-related changes from broader economic pressures and compare roles over consistent periods.
Source: ThorstenMeyerAI.com
