AI Job Creation: LinkedIn Finds 1.3 Million New AI Jobs

AI job creation is real, says LinkedIn, which finds 1.3 million jobs despite a hiring slowdown. AI is not the main cause of weak hiring.

By Cosby Khosa · Published 22 October 2026

Key takeaways

LinkedIn’s analysis says global hiring is nearly 20% below pre-pandemic levels, and 20% to 35% below in advanced economies.
It says AI is not the main cause: outside clinical healthcare, hiring patterns are similar for high and low AI-exposure jobs.
More than 600,000 jobs are tied to AI-enabled data centres, and entry-level hiring is described as back in normal ranges.
The 1.3 million jobs figure comes from the headline and its method is not given, so check the original.

AI job creation is more real than many headlines suggest, according to a LinkedIn analysis covered by Allwork.Space. Its title says that AI has created 1.3 million jobs despite a hiring slowdown. The analysis finds that global hiring sits nearly 20% below pre-pandemic levels, and 20% to 35% below in advanced economies. But it says that AI is not the main cause.

Outside clinical healthcare, it reports, hiring patterns are similar for high and low AI-exposure jobs. More than 600,000 jobs are tied to AI-enabled data centres, and entry-level hiring is described as back in normal ranges. The page shows no date.

This article explains the findings, what they mean for the AI jobs debate and how to use them. It is based on a search extract of a news analysis, so it keeps to what the extract stated. The 1.3 million figure comes from the headline, and the extract did not explain how it was counted. Check LinkedIn’s own research for the method.

What the LinkedIn analysis says about AI job creation

The analysis makes four main points. First, hiring around the world is well below pre-pandemic levels. Second, AI is not the main cause of that weakness. Third, AI is linked to the creation of jobs, including over 600,000 tied to data centres. Fourth, entry-level hiring has returned to normal ranges.

The headline number, 1.3 million jobs created by AI, is eye-catching. It suggests that a large number of new roles exist because of AI, such as AI engineers, data centre workers, data annotators and others. The extract does not say how this figure was calculated, so we cannot judge it. Counting jobs created by a technology is hard, since many roles change gradually and it can be unclear which are new.

The key message is a challenge to a common fear. Many people assume that AI is behind the weak hiring market and is taking jobs. The analysis says that the data do not support that view as the main explanation.

It does not say that AI has no effect on any job. It says that, overall, the hiring slowdown has other causes. This distinction matters, and we should keep it in view.

Read the full analysis for detail, including the sources and method.

Global hiring nearly 20% below pre-pandemic levels

The first finding is that global hiring sits nearly 20% below pre-pandemic levels. In advanced economies, it is 20% to 35% below.

That is a large and lasting fall. Before 2020, hiring was at a certain pace. Now it is considerably slower. The difference means fewer people are moving into new jobs each month.

The gap is wider in advanced economies. This may reflect more mature labour markets, higher interest rates, tighter budgets or a larger share of jobs in sectors that have slowed, such as technology and professional services. The extract does not say.

This links to other reports on OneCareer. LinkedIn’s separate summary says that hiring is 23% below pre-pandemic levels, within the range given here. Indeed’s data describes a low-churn market, with hiring near levels last seen about 11 years ago. Marketplace describes a frozen job market.

For job seekers, the figure explains why a search can feel slow even when unemployment is low. If hiring is a fifth lower than before, there are fewer chances at any time.

It is also a reminder that the problem is not new, and not simply about AI. It has built up since the pandemic boom.

Why AI is not the main cause of the hiring slowdown

The second finding is that AI is not the main cause. The evidence given is that, outside clinical healthcare, hiring patterns are similar for high and low AI-exposure jobs.

Here is the logic. Analysts rate each occupation by how exposed it is to AI, meaning how much of its work could be done or helped by AI tools. If AI were driving the hiring slowdown, you would expect hiring to fall more in high-exposure jobs than in low-exposure ones. The analysis says that this is not what the data show. The patterns are similar.

The exception is clinical healthcare, which behaves differently. The extract does not explain why, and we should not guess. Healthcare has its own pressures and drivers.

This is an important finding, but not a final word. Some effects of AI may be slow to appear, since firms are still learning to use it. Some effects may be hidden, such as cutting future roles without layoffs. And the pattern might change as tools improve.

Other reports hold similar views with caution. Indeed’s June snapshot said that the link between AI exposure and postings appeared to be flipping from job destruction to creation. The theme is consistent: the evidence does not yet show widespread job loss caused by AI.

More than 600,000 jobs tied to AI-enabled data centres

The third finding is that more than 600,000 jobs are tied to AI-enabled data centres. This is a concrete example of AI job creation, and it shows that AI job creation reaches beyond software.

Data centres house the computers that run AI and other digital services. Building and operating them needs many kinds of workers. Construction workers and electricians build them. Engineers design them. Technicians maintain them. Energy workers supply power. Security, logistics and administration staff support the operation.

The figure shows that the AI boom creates real, physical jobs, not only software jobs. This is good news for people in trades and technical roles. It is also a reminder that AI depends on infrastructure.

Marketplace, on OneCareer, notes that AI and data-centre investment offsets other weak spots in the economy. That ties in with this finding. Money flowing into data centres supports employment in construction and related fields.

There are caveats. Data centre jobs are concentrated in certain places, so benefits are uneven. Some jobs are temporary, especially in construction. And the demand depends on continued investment, which could slow.

South African readers may note that data centre investment is also growing in the region, though we have no figures here. Check local sources.

Entry-level hiring is back in normal ranges

The fourth finding is that entry-level hiring is described as back in normal ranges. This contrasts with a widespread narrative that AI is swallowing entry-level jobs.

Another LinkedIn voice on OneCareer, Sue Duke, makes a similar argument in a Fortune interview. She says that the data counter the story that entry-level jobs are being swallowed by AI.

Other sources paint a more mixed picture. Indeed’s analysis shows entry-level postings down 7.5% year on year in May 2026, and a consultancy blog cites entry-level postings down more than 35% since January 2023. These measure different things, such as postings versus hiring, and different time frames.

How can they be reconciled? Postings count advertisements, while hiring counts actual hires. Entry-level postings could be lower while hiring is within normal ranges if, for example, employers fill posts faster or use other channels. Or the sources may define entry level differently. The extract does not give enough detail to settle it.

For graduates, the safest reading is that the market is tough but not collapsing. Competition is strong, and effort is needed. Our guide on how to apply for jobs and get your application noticed can help you stand out. But there is no clear evidence in this analysis that AI has wiped out the first rung.

Keep checking new data as it arrives.

How to square AI job creation with the hiring slowdown

At first sight, it seems odd that AI could create 1.3 million jobs while hiring is down by a fifth. The two ideas fit together once you see the difference between a part and the whole.

Total hiring involves tens of millions of jobs across all sectors. AI-related roles are a small, though fast-growing, slice. They can grow strongly while the rest of the market is weak.

Think of a large lake with a few streams flowing in. The streams may be fast, but the lake can still be falling if other outflows are larger. AI job creation is a stream. The slowdown comes from many other factors that affect the whole lake.

This is why LinkedIn’s Jobs on the Rise list, covered on OneCareer, is led by AI engineer even though hiring is 23% below pre-pandemic levels. Specific demand and general weakness can exist together. Robert Half’s data tells a similar story (see Technology Job Market 2026: Robert Half’s In-Demand Roles).

For workers, the lesson is to look at your own field. If your skills link to AI job creation, you may find opportunities. If they do not, the general slowdown may matter more.

For policymakers, it suggests that helping people move into growing areas, through training, is as important as worrying about job losses.

What AI job creation means for job seekers

For job seekers, the findings on AI job creation offer reassurance and direction.

The reassurance is that, so far, AI is not shown to be the main reason why jobs are hard to find. The hiring slowdown has other causes. That may reduce fear, and help you focus on what you can control.

The direction is to look at where AI job creation is happening and see whether you can connect. Roles tied to AI include engineering, data, consulting, annotation, data centre construction and operations, and security. Many of these need skills that can be learned. Hirebase’s data on in-demand technical skills shows which ones are in demand.

You do not need to become an AI engineer. Many roles that support AI need other abilities. Electricians, project managers, procurement specialists, trainers and customer success staff are all part of the picture.

Build AI literacy in your current field. Learn what AI tools can and cannot do, and use them to improve your work. Employers value people who can use tools well. Paylocity’s guide to upskilling for AI is a useful starting point.
Stay realistic. A hiring rate 20% below normal means competition. Apply with care, build networks and keep learning.
Be wary of promises. Some training providers exaggerate the job prospects of AI courses. Check outcomes, and ask for evidence.

What employers and policymakers can learn

For employers, the message is that AI adoption and hiring need not be opposed. The data suggest that AI creates work as well as changing it.

Plan skills early. If you adopt AI, you will need people who can implement, manage and govern it. Train existing staff where possible, starting with a skills audit (see Workhuman’s skills-gap audit guide and our guide to AI reskilling). Hire for potential, as 57% of tech hiring managers do, according to Robert Half.
Be honest with employees. Fear of job loss is high. Gallup finds that 18% of US employees think their job is likely to be eliminated within five years, rising to 23% where AI is in use. Clear communication about plans reduces anxiety.

For policymakers, the findings point to priorities. Support training for growing roles. Invest in infrastructure, but plan for jobs beyond construction. Monitor data on hiring and AI exposure so that policy responds to evidence, not fear.

Education systems can help too. Teach AI literacy, critical thinking and adaptability. Support lifelong learning.

Finally, avoid complacency. The pattern could change. Keep watching the data, and be ready to respond if AI begins to reduce hiring in specific areas.

How to read this AI jobs analysis with care

Several cautions apply.

First, this is a news analysis of LinkedIn research, not the research itself. Details may be lost or altered. Check the original.

Second, the 1.3 million figure comes from the headline, and the extract did not explain how it was calculated. Measuring AI job creation is difficult, and different methods give different answers.

Third, the page shows no date, so we cannot say how fresh the data is.

Fourth, LinkedIn is a company that sells hiring and learning products. Its research may reflect its own users and interests. That does not make it wrong, but independent studies are valuable.

Fifth, the analysis says that AI is not the main cause of the slowdown. That does not mean AI is harmless. Some workers may be hurt, and effects may grow.

Sixth, the healthcare exception shows that patterns differ by sector.

Finally, this article is a summary of a summary. For decisions that matter, go to primary sources.

Final thoughts

AI job creation is real, according to LinkedIn’s analysis, which reportedly finds 1.3 million jobs despite a hiring slowdown. Global hiring is nearly 20% below pre-pandemic levels, and 20% to 35% below in advanced economies, but AI is not the main cause. Outside clinical healthcare, hiring patterns are similar for high and low AI-exposure jobs. More than 600,000 jobs are tied to AI-enabled data centres, and entry-level hiring is described as back in normal ranges.

For job seekers, the message is to look for where AI creates demand, build relevant skills and stay realistic about competition.

Check the original analysis of AI job creation for method and detail, and watch future data for changes.

Sources and further reading

Allwork.Space, LinkedIn Finds AI Has Created 1.3 Million Jobs Despite a Hiring Slowdown