This article aims to introduce the problem concerning liability imputation of damages caused by Artificial Intelligence (AI) applications in the process of identification of the consumer's profile in Brazil, through the deductive method and the bibliographic research technique. In the first section, the concept and characteristics of AI are identified, mainly the lack of limits on the results it can achieve in strategic marketing applications for identification of the consumer's profile. In the second section of this paper, the assumption of civil liability in the Brazilian Data Protection Law regarding the irregular data processing are analyzed, in a comparative approach to the European analogue regulation (GDPR). The proposed analysis leads to the conclusion that agents must be held responsible regardless of negligence in Brazil, because the essential element of imputation is the irregular processing of personal data, once observed occurring in non-compliance with the protection legislation or frustrating the legitimate expectation of data subjects' security.
