On Tuesday 12 February, MEPs adopted by a comfortable majority (572 votes in favour, 54 against, 45 abstentions) a particularly dense own-initiative report (192 paragraphs over more than 60 pages) calling for a comprehensive European industrial policy on artificial intelligence (AI) and robotics that paves the way for an "ethical" and "human-centred" AI.
"My main objective has been to ensure that this is a positive report that signals our willingness to adopt this technology and provide the correct conditions for it to thrive", said rapporteur Ashley Fox (ECR, UK), urging Parliament the day before the vote: “Let us not stand in the way and be a Parliament of Luddites."
Thus, an AI "made in Europe", a term sanctioned by the European Commission (see EUROPE 12155), should, according to MEPs, be “ethical”, respectful of data protection (particularly in the medical field) and should take into account social issues as well as be subject to a form of legal liability.
This last point heralds major political-philosophical debates. One source cites the case of a fatal accident caused by an application using AI - a self-driving car, for example. Who is responsible? The driver, the company that developed the application or the application itself? A difficult question, to which the Members have not yet given any answers.
In addition, the European Parliament insists on a human-centred approach. According to MEPs, AI and robotics have the most positive impact when they aim to support people at work and at home. This human focus also allows some form of control over AI and robotics.
The issue of job loss caused by the rise of these new technologies is also addressed, with the text emphasising the need to train workers in the future skills needed in the labour market.
Jurisdiction battle. Writing the report was painful, as the rapporteur noted with regret, after several months of battle to determine which committee would have jurisdiction. As time was running out, the issue was only partially resolved, with the ‘Industry, Research and Energy’ Committee being deemed responsible, but with 6 other associated committees. Hence an extremely long resolution, to the great displeasure of a parliamentary source, who regretted the "dissolution of the message" and the weakness of its expression.
As recalled by Environment and Maritime Affairs Commissioner Karmenu Vella, an expert group on AI is expected to deliver ethical recommendations in March. (Original version in French by Pascal Hansens)