We are all overwhelmed with feelings of optimism when we talk about the feats of artificial intelligence. Of course the human being anguished by concerns of safety, health or environmental disasters, he chooses artificial intelligence as a remedy for his anxieties. It is quite clear that artificial intelligence brings huge results to life problems particularly in terms of health, technological progress and in the field of astrophysics but few people are interested in knowing that artificial intelligence can also be used for criminal purposes. Or so many of us prefer to ignore it. This risk of criminal act ranges from hacking of any kind to scams and by intox in the manipulation of information. artificial intelligence in bad hands become a nightmare for the planet in the future. According to Lewis Griffin, a computer science researcher at University College London (UCL), "the expansion of the capabilities of technologies based on artificial intelligence is accompanied an increase in their potential for criminal exploitation.» In his research Lewis Graffin has compiled, of course in collaboration with his colleagues, a list of 20 illegal activities committed by artificial intelligence, and he classified them by order of their degree of potential harm. These are stolen earnings or profits,easy to implement and difficult to spot and stop. In this article we will introduce you to six top activities on the list of Lewis Graffin. Others will be developed in our following publications. in this series of criminal acts committed using artificial intelligence. 'Robot Burglars' Breaking Into Someone's Apartment seem not to be the most dangerous even if the potential to frighten is very big. These robbery robots can be foiled and they can't touch more people at a time. However, false information distributed by "bots" (a bot is just an abbreviation of a robot) have ability to destroy the image of a known personality. They also have a strong ability to blackmail personalities from the world of show business. We call them also "deepfakes". This is to say how dangerous their level of destruction is. In addition what makes even more afraid of these deepfakes is that it is complicated to fight them so that he keeps this high level of being able to cause enormous economic and social harm for someone's life. The most serious threats from artificial intelligence are seven in number. First fake videos. This is the usurpation of the identity of a person and then make him say or do things he has never said or done and all for the purpose of requesting secure data, manipulating public opinion to damage its reputation. These pirate videos are almost impossible to detect. Then the hacking of self-driving cars. Take orders for these vehicles autonomous is even more dangerous for the safety of people. This hack may have the purpose of committing terrorist acts because one can get away with it. serve as a weapon of war. And also a hacked autonomous vehicle can be used to cause accidents. In third place comes Tailored Phishing. It consists of launching personalized and automated messages to build capacity to fish for Internet users in order to enter their databases to have access to the secure information or install cookies or malware. In fourth place comes the hacking of AI-controlled systems. It is disruption of infrastructure, for example by causing a breakdown widespread electricity loss, traffic jams or the total stoppage of operation of food logistics. The fifth criminal activity is Large Scale Blackmail. This method consists of collecting personal data and then sending messages automated threats. Artificial intelligence can also be used to produce false evidence. For example in a trial of a rape charge. And finally in sixth position is -False information written by AI-. Se hiding behind a source that may seem reliable, artificial intelligence can help write propaganda articles. It could also be used for produce many versions of a topic to increase its visibility and credibility.