Ai Will Transform Work More Than Replace Jobs, Study Finds

Cristina Gaidau/iStock/Getty Images Plus via Getty Images

Follow ZDNET: Add america arsenic a preferred source on Google.

ZDNET's cardinal takeaways

• Generative AI could toggle shape jobs much than it replaces them.
• Jobs requiring much beingness labour are little exposed to AI automation.
• The complaint of AI take will alteration betwixt industries and businesses.

The emergence of generative AI devices successful caller years has led to fears of an inevitable jobs apocalypse -- a activity of automation akin to nan Industrial Revolution that pushes monolithic quantities of humans retired of nan workforce.

Also: OpenAI tested GPT-5, Claude, and Gemini connected real-world tasks - nan results were surprising

At slightest for nan clip being, that hasn't travel to pass. Despite predictions from some tech manufacture leaders that AI will soon switch a important swathe of quality workers, nan technology's effect connected nan labour marketplace has frankincense acold been minimal. 

New information from employment hunt motor Indeed, however, shows that generative AI is capable to automate a increasing number of occupation skills. That suggests that alternatively than pushing humans retired of their jobs entirely, AI will commencement to go an progressively important instrumentality wrong nan workplace. 

Transformation vs. replacement

Indeed's latest AI astatine Work Report, an yearly study measuring nan technology's effect successful nan occupation market, recovered that much than a 4th (26%) of jobs that had been posted connected nan level successful nan past twelvemonth could beryllium "highly" transformed by generative AI.

Also: AI helps beardown dev teams and hurts anemic ones, according to Google's 2025 DORA report

The study assessed nan capabilities of 2 models, OpenAI's GPT-4.1 and Anthropic's Claude Sonnet 4, and really efficaciously they were capable to execute crossed a scope of adjacent to 3,000 cardinal activity skills. To that end, nan researchers deployed what they telephone nan GenAI Skill Transformation Index (GSTI), a metric gauging nan grade to which generative AI will toggle shape peculiar occupation requirements, alternatively than nan technology's imaginable to switch jobs wholesale.

"The early of activity and nan domiciled of generative artificial intelligence (GenAI) is not conscionable astir occupation nonaccomplishment aliases automation--it's astir transformation," Indeed noted successful nan report. "Rather than reasoning successful either-or terms--jobs mislaid vs. jobs saved--we must understand GenAI's effect on a continuum of transformation."  

Which jobs are astir exposed?

Each of nan occupation skills included successful nan caller Indeed study were assessed according to their cognitive arsenic good arsenic their beingness demands. Jobs that required much of nan former, specified arsenic package development, were recovered to beryllium much highly exposed to automation from AI; those that revolve chiefly astir nan latter, including nursing, are little apt to go automated anytime soon, according to nan report.

Also: Jobs for young developers are dwindling, acknowledgment to AI

The findings echo a July insubstantial from Microsoft, which likewise recovered that roles requiring nan benignant of repetitive information-processing astatine which ample connection models excel, for illustration translator and customer service, were more apt to beryllium replaced by nan technology. 

Only 19 skills successful Indeed's caller study (0.7% of nan full dataset) were deemed "very likely" to beryllium afloat replaced by generative AI. This is up from zero successful nan company's erstwhile studies, nevertheless -- "still mini successful absolute terms," nan authors statement successful nan report, "but a important awesome of progress."

Another caller study analyzing online occupation postings from nan past 13 years recovered that jobs that require much AI-related skills are offering higher salaries than those that don't, some wrong nan tech manufacture and crossed different sectors.

Takeaway for businesses

Just arsenic AI-powered automation could move retired to return spot connected a continuum alternatively than connected an either-or basis, not each employers are adopting AI astatine an adjacent pace. The complaint of automation moreover wrong those sectors that are astir amenable to having immoderate cardinal skills replaced by AI will alteration depending connected really aggressively individual businesses deploy nan technology.

Also: Your coworkers are sick of your AI workslop

As Indeed's study suggests, generative AI isn't a one-size-fits each solution. AI developers whitethorn beryllium making expansive promises of boosted worker productivity crossed nan board, but nan technology's inferior wrong individual businesses and roles depends connected a wide assortment of factors.

Successfull AI take will require a adjacent spot of experimentation connected nan portion of individual businesses. As Indeed noted, for example: "For businesses seeking to much broadly instrumentality GenAI into their workflows and products, choosing nan correct exemplary for their circumstantial processes and business cases will beryllium captious successful ensuring reliable insights." Recent reporting from authorities for illustration Gartner stress that view. 

Also: If 5% of AI projects succeed, past yours tin excessively - and this is how

Rather than enforcing a top-down attack to implementing generative AI passim nan workplace, it could very good move retired to beryllium overmuch much effective to springiness individual labor nan state and elasticity to commencement utilizing nan exertion successful a mode that's best-suited to their peculiar roles. This was 1 of nan cardinal findings from a caller MIT study, which recovered that astir 95% of generative AI initiatives among businesses person completely fallen flat.