With the expansion of the use of computers and the Internet in everyday life, crimes related to cyber space have also increased. This shows the importance of adapting legal rules and regulations to new technologies and realities. This may cause legal challenges. Accordingly, the purpose of this article is to examine the question of what is the approach of the criminal law of Iran and the European Union to the criminal liability of artificial intelligence? This article is descriptive and analytical and it has investigated the mentioned subject by using the library method. The findings indicate that there is no criminal liability for artificial intelligence in European Union law and the requirement for criminal liability of artificial intelligence is that it has a legal personality. Criminal liability is not defined for artificial intelligence in Iran's criminal law. The impossibility of achieving the goals of punishment, the unenforceability of corporal and deprivation of liberty punishments and financial punishments are the most important challenges of assigning criminal responsibility to artificial intelligence.