In general, AI is about the automation of intelligent behaviour. A distinction is commonly made between so-called “strong” and “weak” AI. In the first alternative, this kind of AI has (at least) the same intellectual abilities as a human. Although such AI is not yet supposed to exist, there is disagreement in the research community only about "when" this will be the case, and not "if" it will occur.

The systems that exist to date are those of “weak” AI, which - typically for very specific application problems - imitate human intelligent behaviour. The recent major successes have been achieved in particular within the framework of machine learning processes in the construction of "intelligent systems".

In the case of new, as yet little-researched technologies, it is obvious that risks should be limited by regulation. However, the inherent risks of a technology only become clear once it has been sufficiently researched and has been in operation for some time. Legislators are currently faced with the difficult situation of finding the right balance between risk limitation and openness to technology. If regulation is set too high, there is a danger of stopping promising development paths too early on the one hand, while on the other hand lax requirements may leave high-risk systems inadequately regulated. What is needed is a balanced approach that harmonises the enabling functions of legislation with regulatory approaches of a passive nature, for example through liability law.

It is clear that we are only on “day 2” at most of a rapid development that has only just begun, which will permanently change numerous other areas of life and which will require customised regulation and handling for each sub-area.

The many legal issues associated with the various areas are referred to in the riders (links).

Artificial Intelligence - Links to further information