Several problems in Cyber ​​Security need to be solved with the help of Artificial Intelligence (AI). As examples, we can mention: detection of attacks in networks and detection of malicious software, correction of vulnerabilities in codes, blocking of unwanted messages (SPAM). In 2019, the first AI algorithm that outperformed human attacks on an encryption algorithm was presented. The machine can create algorithms and detect weaknesses in computer systems better than humans. Despite the many benefits of AI, AI has been used almost exclusively to collect personal data and deliver targeted advertisements for years. With the increasing number of sensors and the volume of data available, AI algorithms are getting better and better, and they can even manipulate public opinion. Even if there is security between the sensors and the service provider, it is essential to guarantee customer privacy. Homomorphic encryption is a solution under development to provide security with privacy. By joining such a technique to AI, we can identify people and patterns, informing only those who are interested. For example, an AI that works with encrypted data from cameras on beaches could provide only the drowning images to save lives and robbery images for prevention, without anyone having to monitor all the captured images. Aiming at these new technologies that arise from the application of AI to Cybersecurity and Privacy, LNCC is offering the discipline of AI applied to Security in its doctoral program and developing projects in this area.