Abstract:

The legal governance of artificial intelligence (AI) technologies is difficult because the intellectual property created by AI is relatively independent of natural persons. Since current AI technologies are still intelligent tools, legislation focuses on the behaviors of the legal subjects behind the AI technologies instead of the AI technologies themselves. In criminal law, it is necessary to know how AI technologies function and how algorithms are applied. Legal subjects related to AI technologies should be held criminally culpable to the extent algorithm security is a factor. When specific implementation schemes are prepared, it is necessary to define the functions and positioning of criminal law in AI governance activities. What should be avoided is unnecessary obsession over proof of technical logic or confusing technical functions in criminal laws with functions in prepositional laws.

Key words: artificial intelligence, technological risk, algorithm security, legal system, criminal law imputation