Goldman Sachs says generative AI could impact 300 million jobs — here’s which ones

As artificial intelligence products like ChatGPT aim to become a part of our everyday lives and we learn more about how powerful they can be, there’s one thing on everyone’s mind — how AI could impact jobs.

“Significant disruption” could be on the horizon for the labour market, a new Goldman Sachs report dated said. The bank’s analysis of jobs in the US and Europe shows that two-thirds of jobs could be automated at least to some degree.

In the US, “of those occupations which are exposed, most have a significant — but partial — share of their workload (25-50 per cent) that can be replaced,” Goldman Sachs analysts said in the research paper.

Around the world, as many as 300 million jobs could be affected, the report says. Changes to labour markets are therefore likely – although historically, technological progress doesn’t just make jobs redundant, it also creates new ones.

The use of AI technology could also boost labour productivity growth and boost global GDP by as much as 7 per cent over time, Goldman Sachs’ report noted.

Jobs most and least affected by AI automation

Certain jobs will be more impacted than others, the report explains. Jobs that require a lot of physical work are, for example, less likely to be significantly affected.

In the US, office and administrative support jobs have the highest proportion of tasks that could be automated with 46 per cent, followed by 44 per cent for legal work and 37 per cent for tasks within architecture and engineering.

The life, physical and social sciences sector follows closely with 36 per cent, and business and financial operations round out the top five with 35 per cent.

On the other end of the scale, just one per cent of tasks in the building and ground cleanings and maintenance sector are vulnerable to automation. Installation, maintenance, and repair work is the second least affected industry with 4 per cent of work potentially being affected, and construction and extraction comes third from the bottom with 6 per cent.

Data for Europe is slightly broader, but paints a similar picture with clerical support roles being most affected as 45 per cent of their work could be automated, and just 4 per cent of work in the crafts and related trades sector being vulnerable.

Overall, 24 per cent of work in Europe could be automated — just below the 25 per cent average in the U.S.

The countries most affected

These figures shift when looking at automation through AI on a global scale.

“Our estimates intuitively suggest that fewer jobs in EMs [emerging markets] are exposed to automation than in DMs [developed markets], but that 18 per cent of work globally could be automated by AI on an employment-weighted basis,” the Goldman Sachs report said.

According to the bank’s analysis, Hong Kong, Israel, Japan, Sweden and the US are likely to be the top five most affected countries. Meanwhile, employees in mainland China, Nigeria, Vietnam, Kenya and, in last place, India, are the least likely to see their work being taken over by AI technology.

But while the data shows that AI will undoubtedly impact the labour market, it’s not yet clear how disruptive it will really be, the report concludes.

“The impact of AI will ultimately depend on its capability and adoption timeline,” it says, adding that two key factors will be how powerful AI technology really becomes and how much it is used in practice.

CNBC