Liability for (semi)autonomous systems: robots and algorithms

摘要

Liability for robots and algorithms is at present unclear. A comparative overview shows that while several grounds for liability may apply, depending on the jurisdiction, there are still significant gaps, in particular regarding liability for algorithms. Several changes would be required to provide effective protection of interests that may be harmed by defective autonomous systems. By careful regulation a proper balance may be obtained between allowing innovation without undue harm for society.