AI has not yet led to the loss of many jobs in OECD countries, but rules and a map of risks are needed, especially those that escape the rules already adopted, to manage it in a way that produces positive effects on businesses and workers. “We made estimates, asking entrepreneurs and workers about the impact of artificial intelligence within the company,” explains Stefano Scarpetta, director for employment, work and social affairs at the OECD, on the sidelines of the G7 Labour which began today in Cagliari. “In reality, the data that emerge is that, at least to date, there has been no replacement of workers with AI models.
Employment has not decreased. What will change is the type of work tasks that workers will perform within many companies in all sectors. And so we need to invest in the skills of these workers so that they remain complementary to what generative AI models can do today and will be able to do tomorrow.” “I hope that after the summit, the Ministers of Labor will adopt an action plan for the use of AI that is reliable,” said Scarpetta, one of the delegates present at the meeting chaired by the Minister of Labor, Marina Calderone.
“This is a fundamental point. There are already guidelines on how to develop AI models, but we need to provide guidelines and good policies for a use that has benefits for companies and also for workers and that reduces the potential risks associated with this technology.”
Scarpetta focuses on the impact of AI on the labor market, a central theme of today’s meeting. “In reality, there will be few jobs that will disappear completely, but many will be transformed, even radically,” the OECD delegate predicts. “It is interesting how generative AI is more in competition with highly qualified workers. It performs non-routine tasks and also with cognitive skills. However, what we see in the data is that, while workers with high levels of skills can be complementary with AI, even generative AI, perhaps those who will suffer the most from a reduction in the type of jobs, quality and remuneration are workers with lower levels of qualifications. It is in them that we need to invest.” Scarpetta sees “great potential but also risks” in AI.
“In our countries,” he recalls, “we already have a very broad regulatory framework that protects against privacy, discrimination and other issues. The problem is to understand whether AI needs to be regulated more than the regulatory framework we already have because, in some way, it gets in the way of some loopholes in the regulatory systems. So, we need to identify the risks. Map these risks and understand which areas to intervene in. Well, the action plan on AI that the labor ministers are discussing does just that: it identifies the risk areas, the priorities in terms of policy interventions.”