Warning against the emergence of a machine-dominated dystopia, as anticipated by Edward Morgan Forster at the beginning of the 20th century, the European Agency for Safety and Health at Work (EU-OSHA) has made a series of recommendations to better regulate artificial intelligence in two recent articles.
The European Agency notes the increasing use of ‘big data’ and artificial intelligence in the workplace and is concerned about the risks that these new technologies pose to the health and well-being of workers. These risks are numerous: risk of unfair treatment by management and human resources in their AI-assisted decisions, endangering workers, discrimination, massive elimination of low-skilled jobs, an increase in accidents involving robots and workers, breaches of privacy... According to the article, China’s citizen rating system gives a foretaste of the ethical abuses that these technologies can lead to.
However, the organization also sees beneficial effects, such as the ability to reduce “human bias” in the decision-making process with appropriate algorithms, by identifying discriminatory risks, for example. In addition, AI can free up people's working time by drastically increasing their productivity. The key, according to the OSHA article, is therefore to set up an “assistive and collaborative” AI instead of a “universal AI”. The organisation considers that workers must be constantly consulted. Finally, the development of these technologies must take the human being as a starting point - in this case, the people in the work environment.
Artificial intelligence is becoming increasingly important in the European political arena: several European political groups want to strengthen the European framework in this area (see EUROPE 12250/8), and the European Commission wants to develop AI that is “Made in Europe” (see EUROPE 12231/10).
To consult the articles, see here: https://bit.ly/2VE3JsW and here: https://bit.ly/2HpbxVi. (Original version in French by Pascal Hansens)