Robots, ethics and law: where have we got so far?


Are robotic innovations regulated? What happens if a robot makes a lethal mistake in the operating room? And what if a driverless car causes an accident? Who is to blame? The issue, one of the hottest at the moment, is at the top of the European Parliament’s agenda, although it is still far from being resolved (we discussed this issue also here). European legislators’ interest is due both to the importance of this new market in economic terms and to the social impact of these technologies: from increased road safety to progress in surgery to a better life for the disabled. In this scenario, the debate in Italy is ongoing as a result of the progress made by Robolaw, an international research project funded precisely by the European Parliament and coordinated by Istituto Superiore Sant’Anna in Pisa.

Having defined the boundaries of an incredibly large universe – from drones to bionic prostheses, from exoskeletons to ‘non-corporeal’ artificial intelligence made of pure software – the project intends to carry out a systematic work on the ethical, social and  legal issues arising from robotics.

What are the legally relevant features of robots? According to the Robolaw team, not only the ability to perceive external data and analyse them to proceed to an action, but also the ability to learn (machine learning) and reproduce human behaviour, mimicking the human decision-making process, a feature that can often cause behaviour that is not even foreseen by robot programmers.

Yet, robotics is a sector that need not be regulated from scratch: many existing laws can operate effectively, also for artificial intelligence. Steps need
to be taken only where the legal system proves to be lacking. The resolution approved by the European Parliament on 16 February goes precisely in this direction, for it contains a number of civil law recommendations on robotics.

Alongside traditional directives, the resolution envisages responsible self- regulation tools – licences for robot programmers and users, codes of conduct
for researchers and scientific committees, regulation by design, a sort of binding rules embedded in the machine’s design.

In this last respect, a deeper consideration might arise which, far from being resolved, significantly distances men from machines: that of being able to
consciously choose, as beings having morals, to comply with those rules. In this detail lies the exclusively human discernment between good and evil.