EU – Commission puts AI regulation proposal up for discussion27. April 2021
EU – Commission puts AI regulation proposal up for discussion
The European Commission has now published a regulatory proposal for the use of AI. The Artificial Intelligence Act sets limits on the use of AI in numerous areas . These include: self-driving cars, job allocation, credit allocation and the assessment of exams. It further regulates the use of AI by law enforcement and court systems.
Prohibitions are also included. These include, for example, live facial recognition in public spaces, with exceptions for national security and some other purposes.
“In the field of artificial intelligence, trust is a must, not a nice-to-have,” said Margrethe Vestager, the European Commission vice president in charge of digital policy for the 27 nations. “With these groundbreaking rules, the EU is leading the way in developing new global standards to ensure AI can be trusted.”
The 108-page draft will cause significant headaches, especially for major tech companies such as Google, Facebook, Microsoft and Amazon, all of which have invested significantly in AI development. Violations of the new regulations could result in fines of up to 6% of global sales.
This plan is not exactly being received with enthusiasm there. “The question that every company in Silicon Valley will be asking today is: should we remove Europe from our maps or not?” said Andre Franca, director of applied data science at CausaLens, a U.K.-based AI startup, in an interview with Fortune.
The law is “a damaging blow to the Commission’s goal of making the EU a global AI leader,” Benjamin Mueller, a senior policy analyst at the Center for Data Innovation in Washington, DC, a consulting firm indirectly funded by U.S. tech companies, told Fortune. The proposed law would create “a thicket of new rules that will cripple companies hoping to develop and use AI in Europe,” leaving European companies behind, he suggested.
Herbert Swaniker, a lawyer at Clifford Chance, an international law firm based in London, said the proposed law “will require a fundamental shift in the way AI is developed” for providers of AI products and services.
But criticism is not coming from the U.S. alone. Media law expert Stephan Dreyer of the Hans Bredow Institute at the University of Hamburg, for example, says the scope is “massively unclear.” The term AI is so broad “that almost any software currently in use can fall under it.” It is open, he said, to what extent operating systems that include AI-based software such as voice assistants would fall under the regulation. Dreyer fears that this could lead to a “high degree of legal uncertainty for companies” and thus to disadvantages in international competition.
Christiane Wendehorst, a professor of civil law at the University of Vienna, takes a completely different view. She believes the draft addresses “a number of problems without endangering Europe as a technology location.” In her view, it “finally formulates red lines that must not be crossed by AI applications, such as certain forms of targeted manipulation.”
These are the first reactions to this EU initiative. We will keep you informed about further activities.