인공지능(AI)의 민사책임 - EU의 입법을 중심으로 -

Abstract

In the digital age, the European Commission has published the AILD(Artificial Intelligence Liability Directive). The Commission proposed to supplement and modernize the EU's non-contractual civil liability regime, introducing new special rules for damages caused by AI systems. The new regulation is intended to ensure that victims of AI systems receive the same level of protection as victims of other technologies in the EU. The AILD reduces the burden of proof of damage caused by AI systems by creating a 'presumption of causal relationship' that can be disproved. It also gives courts the power to order the release of evidence about high-risk AI systems that appear to have caused harm. And, the European Commission announced the revision of the PLD(Product Liability Directive). The revised PLD aims to transform the EUs product liability regime to suit the digital age. The amendment introduces new regulations addressing liability for products such as software and digital services containing AI systems. In addition, the victim's burden of proof is alleviated by the presumption of defects and causal relationships, and the disclosure of evidence by the courts. Meanwhile, Korean laws that apply to artificial intelligence include the Tort Act and the Product Liability Act, both of which do not yet have special regulations on AI. In this situation, the EU's AILD(Artificial Intelligence Liability Directive) and the revision to the PLD(Product Liability Directive) give us many implications.