Two headlines, two stories. United Kingdom postings mentioning AI run 127% above their pre-pandemic baseline, while total postings sit 27% below it. In the United States, postings hold at baseline while AI leads all reasons for announced cuts. Which do you believe?
Both stories can be true. We covered the AI hiring boom and the core hiring paradox: heavy adopters hire more, not less. Here we pull the two markets apart and hand you a three-question test for judging which findings apply to your business.
Why is UK AI hiring surging while overall UK job postings sit 27% below pre-pandemic levels?
The surge is compositional: postings mentioning AI ran 127% above their February 2020 baseline while total postings sat 27% below it, hitting a record 9.4% share by end-June 2026, up from 7.5% in February.
That index is seasonally adjusted, a seven-day trailing average with 1 February 2020 = 100, so postings sit at roughly 73.
AI mention analysis is a keyword scan for Machine Learning, Artificial Intelligence, LLMs and ChatGPT; a mention can signal deep ML expertise or just a screening notice. It captures employer expectations and ad composition.
Data and analytics postings lead at 47% AI mentions, software development at 41%; care, cleaning and driving sit below 1%. In marketing, HR and accounting, overall postings have fallen even as AI mentions more than doubled.
The backdrop is soft: Office for National Statistics (ONS) vacancies hit 707,000 in May to July 2026, the lowest outside the pandemic since late 2014, and posted wage growth cooled to 3.9%. Indeed Hiring Lab calls the UK a low-hire, low-fire market, while the Bank of England reads the redundancy rate as elevated relative to recent years. Different measures and windows, but both consistent with a frozen market, and with no sign of AI-driven replacement at scale.
The raw data is on the Indeed Hiring Lab Data Portal and Hiring Lab’s GitHub. Caveats: a mention is not an AI role; repostings and sector mix shift the share; regional variation is wide. Junior hiring is weak too: entry-level roles down 14% in the year to April 2026. See the entry-level hiring decline.
US vs UK: why have AI hiring trends diverged so sharply between the two markets?
Across the Atlantic, the picture reads differently. Two regimes, two instruments. UK postings sit 27% below baseline with AI mentions at records; US postings sit at 101.0. The split is macro conditions and market structure, not a single AI effect.
In the US, new postings averaged 100.0 in the first half of 2026, and announced hiring plans ran 107,500 year to date, up 25% on a year earlier. Alongside that, 112,713 AI-attributed cuts kept AI the leading reason for announced reductions for a fifth consecutive month. US posted wages rose 2.4% over the year, against 3.9% in the UK.
No US counterpart to the UK AI-mention series exists; the US side gets read through Challenger, Gray & Christmas layoff attribution instead, mixing a “Technological Update (possibly AI)” category with disputed cases like Montefiore. Attribution is not causation, and the two signals are not comparable; see our statistics guide.
What drives the split? Macro conditions first: the Bank of England held Bank Rate at 3.75% in April, and US rate-cut expectations have all but disappeared. The energy shock hits the UK hardest: electricity averaged US$111.65 per megawatt in May against US$28 in the US. Sector mix and adoption depth follow: about 17.5% of US businesses use AI in at least one function, yet 90% of surveyed firms report no employment or productivity impact over the prior three years. US churn and heavy-adopter hiring, the core paradox, contrast with the UK’s low-hire, low-fire labour market.
Different instruments and structures mean the UK mention share and the US attribution count both pick up composition and intent as much as AI demand; neither market is ahead. The pattern holds across global AI hiring patterns.
Australia sits between the two. The Department of Workplace Relations (DWR) finds no large-scale AI job loss: software development employment is up 25% since November 2022, and AI-exposed occupations grew 5.6% against 9.5%. PwC Australia puts sector wage premia at 59% for tech, media and telecoms, 57% manufacturing, 43% financial services, 42% consumer markets. Watch product-cycle demand and energy-linked inflation, not US layoff attribution or metro concentration.
How can you judge whether macro AI-hiring findings apply to a 50-500 employee SaaS company?
Test every finding with three questions: what does the measure count, which market was it built in, and is it a condition or a cause? Talent-pool economics, wage premia and the junior-market reset travel to your firm; headcount causality does not.
No source segments these findings by company size, and your firm differs from the aggregate on sector, sales motion and hiring mix, so treat any aggregate claim as unverified at your scale until you test it.
Conditions travel: knowledge sectors have run structurally looser than presence sectors since 2022, professionalised jobs grow twice as fast as democratised ones with 42% faster wage growth, and AI-exposed junior roles are roughly seven times more likely to demand senior skills.
Causality does not travel: headcount changes at any single firm, US metro concentration in New York, San Francisco and Seattle, and AI-attributed cuts are not signals your business should act on.
So run the test.
- What does the measure count? A mention is not a role; an AI-attributed cut is an announcement, not causation.
- Which market was it built in? UK-only postings and US attribution data do not map onto a firm in a different market and size band.
- Condition or cause? If it describes a market condition, import it; if it asserts a cause, do not.
Reapply the core paradox and verification framework to your own postings, churn and wage bill, and benchmark salaries against your sector’s premium, not the headline. See which tech roles are actually growing for role-mix planning; the graduate pipeline is where the junior reset hits first.
The US-UK split is market structure and macro conditions, not a single AI effect; recomposition, not replacement, is the throughline of the series. Segment every headline by market, measure, company size and role mix before acting. Import conditions, never conclusions. Treat US and UK data as leading indicators, checked against Australian DWR and PwC Australia evidence.
Frequently Asked Questions
Where can I find the raw job postings data behind Indeed’s UK AI hiring analysis?
The underlying postings data is on the Indeed Hiring Lab Data Portal, with chart data also published on Hiring Lab’s GitHub repository as downloadable CSV files, typically refreshed a day or so after each analysis goes live. To verify the 27% below-baseline figure, cross-check ONS vacancy and employment tables and the Bank of England’s Monetary Policy Report. The series covers UK postings only, so treat it as an advertising measure, not an employment count.
What does “27% below pre-pandemic” actually mean?
It compares advertised UK job postings with their level on 1 February 2020, Indeed’s baseline of 100, using a seasonally adjusted seven-day trailing average. A 27% shortfall means postings sit at roughly 73 on that index; it is not a 27% fall in employment or vacancies. ONS vacancies, a separate measure, were 707,000 in May to July 2026, the lowest outside the pandemic since late 2014.
What does “low-hire, low-fire” mean for tech hiring?
It describes a frozen labour market where employers neither hire aggressively nor cut jobs at scale, and Indeed uses the phrase to characterise the UK. For tech leaders, the practical effect is that fewer people move roles, so competition per vacancy stays high while hiring plans stay cautious. The pressure lands unevenly: UK entry-level hiring was down 14% year over year by April 2026, with graduate postings at their lowest for the season since 2020.
Does an AI mention in a UK job posting mean the role is an AI job?
No. AI mention analysis is a keyword scan for terms such as artificial intelligence, machine learning, LLMs and ChatGPT, so it captures employers’ expectations, including screening notices, rather than dedicated AI roles. The mentions concentrate where AI is a job requirement: data and analytics postings sit at 47% and software development at 41%, while in-person categories such as care and cleaning remain below 1%.
Why are AI mentions concentrated in data and analytics and software development?
Those are the roles closest to building and running AI systems, so employers there advertise AI proficiency most heavily: 47% of data and analytics postings and 41% of software development postings mention it, against below 1% in beauty, care, cleaning and driving. The more significant signal is that AI language is spreading into marketing, HR and finance postings even as overall hiring in those categories declines.
Is it true that AI is replacing tech jobs in the US?
The US data tracks attributions, not proven causation. Challenger, Gray & Christmas counts 112,713 announced cuts citing AI this year through July, about 24% of all announced cuts, and its method includes a “Technological Update (possibly AI)” category and disputed cases such as Montefiore. Announced hiring plans rose 25% year on year over the same period, and the heaviest AI adopters are still adding headcount.
Are UK tech salaries rising because of the AI hiring surge?
Not across the board. UK posted wage growth has cooled to 3.9% annually, its lowest since early 2022, so the record AI mention share reflects posting composition more than pay inflation. Where a premium exists it is concentrated: PwC finds jobs professionalised by AI growing twice as fast as democratised ones, with 42% faster wage growth, and UK pay pressure remains strongest in specific tech, healthcare and engineering roles.
Are UK companies adopting AI faster than US companies?
No clear evidence supports that claim, and the UK posting surge is not an adoption measure. The survey evidence points the other way: about 17.5% of US businesses used AI in at least one function in early 2026, and roughly 70% of surveyed large firms report actively using it, even as about 90% report no employment or productivity impact yet. The UK signal counts intent written into job ads, not deployment.
What should I watch to know whether the US and UK trends are converging?
Track the drivers, because the split reflects market conditions rather than one AI effect. Watch UK job postings and ONS vacancies, UK posted wage growth at 3.9% and cooling, and the rate path after Bank Rate was held at 3.75% in April 2026. On the US side, watch whether AI-attributed cuts keep leading announced reductions while hiring plans stay up 25% year on year.
Why hasn’t Australia seen the same AI job losses as the US?
The local evidence shows recomposition, not displacement. Australia’s DWR report finds no large-scale AI job loss, with software development employment up 25% since November 2022, and AI-exposed occupations growing 5.6% against 9.5% for the least exposed. That slower growth is a gradual shift rather than a US-style cut cycle, so treat US layoff attribution and metro concentration as context, not a template.
