Will AI create or take away more job opportunities?

AI's net effect on jobs is still hotly debated — I'll pull the latest research on what's being created versus displaced.

AI will do both - take away a lot of jobs and create a lot of jobs - but most major forecasts expect it to create slightly more than it destroys in the long run, with a painful transition in the short to medium term.

The consensus of the big forecasts: Net gain, but huge churn

1. World Economic Forum - Net +78 million by 2030
The most comprehensive global employer survey is the WEF Future of Jobs Report 2025. Its 2025 edition surveyed 1,000+ employers across 55 economies and found:

> 92 million jobs displaced, but 170 million created — a net gain of 78 million

The WEF is the only major forecaster to properly account for job creation alongside destruction. It estimates about 22% of today's jobs will be remade by 2030.

2. McKinsey Global Institute - Also net positive, after a wave of displacement
> AI and automation will cut demand for about 36 million U.S. jobs by 2035 while growth elsewhere creates about 41 million

In a more recent report McKinsey said automation could reduce labor demand in the U.S. by 21% — roughly 36 million jobs — over the next decade.

McKinsey notes, like other technological revolutions before it:

> the age of AI may ultimately create more jobs than it destroys

But it warns this will still require "the largest and most sustained workforce transformation in US history" and about 11 million workers — about 7% of all US employees — may need to find new professions by 2035.

3. Goldman Sachs - Exposed is not the same as lost
Goldman's famous 2023 headline was:

> 300M jobs exposed globally; 25% of work tasks automatable

But the bank stresses "exposed doesn't mean displaced". The full forecast is:
> Goldman estimates 7% of US workers will actually lose their jobs, while 63% are augmented

Goldman projects 6-7% of the US workforce displaced during the full AI adoption transition, not all by 2030 specifically, and calls the effect temporary as new jobs emerge.

Why the short term feels like more jobs lost

A more granular, real-time analysis from Goldman Sachs in April 2026 looked at actual U.S. payroll data, not forecasts:

> AI substitution — where machines replace human workers outright — eliminated roughly 25,000 jobs per month. AI augmentation — where the technology makes existing workers more productive — created approximately 9,000

> 16,000 net US jobs erased every single month for the past year

Who is absorbing the damage? Goldman found:

> It isn’t seasoned professionals or senior managers. It’s Gen Z and entry-level workers, disproportionately concentrated in the exact administrative, data entry, and customer service roles that AI handles cheapest and fastest

Other studies agree: 41% of employers surveyed by the WEF are planning AI-related headcount reductions even as they plan to create new AI roles.

History gives context

Historically, technology has been a net job creator, not destroyer:

> while new machines will disrupt the world of work, overall, it will create more new jobs than those it...

For example:
> the introduction of personal computers from the 1980s destroyed an estimated 3.5 million jobs in the US (for instance, among typists), but created 19 million new jobs right across the economy

A UK report looking at 144 years found:

> technology has created more jobs than it has destroyed

And researchers note:

> Efficiency gains have historically created more jobs than they destroyed—not by preserving old work, but by spawning entirely new industries

So, create or take away?

The key variable is not the technology itself, but transition: reskilling, mobility, and how fast new roles emerge to absorb displaced workers. Most economists agree AI will produce prosperity - a ~7% GDP boost over a decade by Goldman's estimate - but the gains will be uneven unless retraining keeps up.