Legislation on AI systems

What legislation specifically applies to AI systems?

Currently, there are only a few laws regulating some applications of AI. Those are addressed here.

The moment an AI uses personal data or makes predictions about people that have a significant effect on them, it falls under the scope of the GDPR (for most organisations) and the Dutch Police Data Act (Wet Politiegegevens) when this is done by law enforcement. To process personal data, there must be a processing ground and training or using a model is a form of processing. Furthermore, a data subject may have the right to receive an explanation of how a model works if the outcome of the model has legal consequences or a significant effect on the data subject's life.

In addition to regulating some models, every AI system is also (part of) a piece of software, so the laws that govern software also apply. For example, software copyright law also applies to AI systems. There are currently stricter rules for physical products than for software that is purely digital. If an AI model is incorporated into a product such as a robot or a self-driving car, that AI model is subject to the same rules as the product it is in.