Will AI create or take away more job opportunities?

Will AI be a net job creator or eliminator? I'll pull the latest research to give you a balanced answer.

On an aggregate, global level, the major forecasts say: AI will create more jobs than it destroys - but that headline hides a very uneven and disruptive churn.

The most-quoted forecast: Net +78 million by 2030

The World Economic Forum's Future of Jobs Report 2025 is the most cited projection:

> By 2030, AI and related technologies are expected to create about 170 million new roles while displacing roughly 92 million existing ones, leaving a net positive of 78 million jobs

The report anticipates that by 2030, AI and other information processing technologies will transform 86% of businesses, sparking the creation of 170 million new roles worldwide while making 92 million existing jobs redundant

Details from the same report:

> 170 million new jobs to be created globally by 2030, while 92 million existing roles face displacement

And that is about structural churn, not just AI. The WEF's underlying projection for structural labour market transformation in 2030 is that 22% of jobs today will be created or destroyed between now and then. Of that 22%, 14% of jobs will be newly created, and 8% will be destroyed.

If you isolate just AI and information processing technology, the net gain is much smaller:

> AI and data processing are expected to create 11 million new roles by 2030, while displacing 9 million - a net gain of 2 million jobs in that category alone

Other institutions give a similar net-positive picture, though numbers vary:

The IMF's May 2024 research nuanced that: about 40 per cent of jobs worldwide to be augmented, downgraded or eliminated by AI over time, with exposure at roughly 60 per cent of jobs in advanced economies, 40 per cent in emerging markets and 26 per cent in low-income countries. That research also found about half of the exposed jobs in rich economies could benefit from AI rather than lose out.

McKinsey adds the technical potential layer: 57% of current U.S. work hours could in theory be automated by technology that already exists - which it calls technical potential, "not a forecast of job losses."

Who loses vs. who gains - they are not the same people

The reports agree the pain is not evenly distributed:

Most at risk - routine, predictable, rules-based digital work:
> Data processing and clerical and administrative positions like data entry clerks, cashiers and telemarketers-occupations composed, by definition, of predictable, rote, rules-based work-have been identified, time after time, as one of the fastest shrinking categories

More recent breakdowns:

Fastest growing - not just coders:

> Farmworkers and delivery workers and laborers and construction workers and salesmen and food processing workers the 'front-line workers'... AND nursing jobs and counselors and social work jobs and personal care aides, along with secondary and university-level education jobs, will represent significant future growth occupations

Technology roles and positions supporting the 'green transition' — like renewable energy engineers — are also among those expected to see the fastest growth, plus AI-specific roles like data annotators, AI trainers, and machine-learning engineers.

Why "net creation" does not mean painless

All the forecasts emphasize three catches:

1. Skills churn is huge. 39% of existing skill sets will become outdated between 2025-2030 and 39% of core skills will likely change significantly over the next six years.

2. The bottom rung is shrinking. 40% of the respondents to the survey behind the report state that they anticipate actually cutting headcount as a result of an ability to have tasks performed by AIs. The IMF warns AI was already hollowing out demand for routine, repetitive work and Singapore's President noted the first rung in employment, your first job upon graduation, is now seeing less demand.

3. Geography and timing. A positive outlook from the top of the economic mountain does not automatically translate to an equitable distribution of gains to the people on the bottom slopes. The fact that there will be 78 million more jobs globally does not provide comfort if new opportunities exist elsewhere.

> By the close of this decade, the story about AI and employment will be about churn, a massive and turbulent but uneven relocation, rather than annihilation

So the answer to "create or take away more?"

Policy advice across the reports is consistent: 85% of employers plan to prioritise workforce upskilling, but success depends on whether retraining, entry-level redesign, and regional support happen fast enough to match the pace of automation.

> A more honest public conversation... would focus rather less on arguing whether the total number of jobs is likely to go down, or up (all evidence... is now leaning toward net gains), and instead rather more on those more granular questions about how to prevent that aggregate transition from appearing... unnecessarily cruel