AI is unlikely to destroy your job unless you’re doing administrative work — about a quarter of such tasks could be lost to the tech.
Artificial intelligence is unlikely to destroy jobs, per the International Labour Organization.
However, in high-income countries, 21 million jobs held by women have the potential to be automated.
3% of jobs in high-income countries held by men face the potential of being automated.
Amid growing fears across sectors of being replaced by AI, a new study by the International Labour Organization, or ILO, published on Monday finds admin workers and women face a greater risk of their jobs being replaced by the tech.
The effects of automation are “highly gendered, with more than double the share of women potentially affected by automation,” per the study.
In high-income countries, 7.8% of jobs held by women have the potential to be automated — which the ILO estimates to be around 21 million jobs.
In contrast, 2.9% of jobs in high-income countries held by men — or around 9 million jobs — face the potential of being automated, as shown in the graf below.
International Labour Organization graph on jobs with automation potential
International Labour Organization
Clerical support workers also face the greatest risk of being affected by AI. About a quarter of clerical workers’ tasks face a high exposure to generative AI, and 58% of their tasks face a medium exposure to the technology.
The study listed jobs such as typists, travel consultants, scribes, contact center information clerks, bank tellers, and survey and market research interviewers that could eventually be automated.
In general, however, most workers’ tasks are not majorly at risk from AI, the study found.
For example, the study found that only 4% of service and sales workers’ tasks face high exposure to generative AI, and 18% of their tasks face medium exposure. “Therefore, the greatest impact of this technology is likely to not be job destruction but rather the potential changes to the quality of jobs, notably work intensity and autonomy,” wrote the study.
The study used GPT-4, the large language model behind ChatGPT, to predict how jobs might be automated.
The authors also noted that AI’s “socioeconomic impacts will largely depend on how its diffusion is managed” and the costs to affected workers could be “brutal.”
This study adds to past research on how AI could disrupt jobs.
A July study by McKinsey also found that generative AI could benefit white collar in the long run instead of replacing them, while a Goldman Sachs study in March found that administrative staff were at particular risk from the technology.