Artificial Intelligence is more likely to augment jobs than to destroy them, a UN study indicated on Monday, at a time of growing anxiety over the potential impact of the technology.

The launch in November of the generative AI platform ChatGPT, which is capable of handling complex tasks on command, was seen as a tech landmark foreshadowing a potentially dramatic transformation of the workplace.

But a fresh study from the United Nations’ International Labour Organization (ILO) examining the potential effect of that and other platforms on job quantity and quality suggests that most jobs and industries are only partially exposed to automation.

Most are “more likely to be complemented rather than substituted by the latest wave of Generative AI, such as ChatGPT”, the ILO said.

“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.”

The study meanwhile highlighted that the effects of technology would vary greatly between professions and regions, while it warned women were more likely than men to see their jobs affected.

It found that clerical work was the category of jobs with the greatest technological exposure, with nearly a quarter of tasks considered highly exposed and more than half of tasks having medium-level exposure.

In other occupational groups, including managers and technicians, only a small share of tasks was found to be highly exposed, while around a quarter had medium exposure levels, the ILO said.

The analysis meanwhile indicated that higher-income countries would experience the greatest effects from automation due to the important share of clerical and para-professional jobs in the job distribution there.

It found that a full 5.5 percent of total employment in high-income countries was potentially exposed to the automating effects of generative AI, whereas only 0.4 percent of employment in low-income countries was.

At the same time, the study found that the share of employment potentially affected by automation was more than twice as high for women as for men, due to women’s overrepresentation in clerical work, especially in high- and middle-income countries

While Monday’s report showed significant differences in the potential impact on AI-generated job losses between wealthy and poorer countries, it found that the potential for augmentation was nearly equal across countries.

This suggests that “with the right policies in place, this new wave of technological transformation could offer important benefits for developing countries”, the ILO said.

It cautioned though that while augmentation could indicate positive developments, like automating routine tasks to free up time for more engaging work, “it can also be implemented in a way that limits workers’ agency or accelerates work intensity”.

Countries should therefore design policies to support an “orderly, fair and consultative” shift, the report authors said, stressing that “outcomes of the technological transition are not pre-determined”.

Artificial Intelligence is more likely to augment jobs than to destroy them, a UN study indicated on Monday, at a time of growing anxiety over the potential impact of the technology.

The launch in November of the generative AI platform ChatGPT, which is capable of handling complex tasks on command, was seen as a tech landmark foreshadowing a potentially dramatic transformation of the workplace.

But a fresh study from the United Nations’ International Labour Organization (ILO) examining the potential effect of that and other platforms on job quantity and quality suggests that most jobs and industries are only partially exposed to automation.

Most are “more likely to be complemented rather than substituted by the latest wave of Generative AI, such as ChatGPT”, the ILO said.

“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.”

The study meanwhile highlighted that the effects of technology would vary greatly between professions and regions, while it warned women were more likely than men to see their jobs affected.

It found that clerical work was the category of jobs with the greatest technological exposure, with nearly a quarter of tasks considered highly exposed and more than half of tasks having medium-level exposure.

In other occupational groups, including managers and technicians, only a small share of tasks was found to be highly exposed, while around a quarter had medium exposure levels, the ILO said.

The analysis meanwhile indicated that higher-income countries would experience the greatest effects from automation due to the important share of clerical and para-professional jobs in the job distribution there.

It found that a full 5.5 percent of total employment in high-income countries was potentially exposed to the automating effects of generative AI, whereas only 0.4 percent of employment in low-income countries was.

At the same time, the study found that the share of employment potentially affected by automation was more than twice as high for women as for men, due to women’s overrepresentation in clerical work, especially in high- and middle-income countries

While Monday’s report showed significant differences in the potential impact on AI-generated job losses between wealthy and poorer countries, it found that the potential for augmentation was nearly equal across countries.

This suggests that “with the right policies in place, this new wave of technological transformation could offer important benefits for developing countries”, the ILO said.

It cautioned though that while augmentation could indicate positive developments, like automating routine tasks to free up time for more engaging work, “it can also be implemented in a way that limits workers’ agency or accelerates work intensity”.

Countries should therefore design policies to support an “orderly, fair and consultative” shift, the report authors said, stressing that “outcomes of the technological transition are not pre-determined”.