Security concerns, including data manipulation, monitoring and Denial-of-Service represent a significant risk to smart grid systems, essential to the power sector. Deep learning-based solutions have become more popular for identifying cyberattacks in different computer environments, and multiple steps have been taken to safeguard diverse systems. Technologies like Generative Adversarial Networks have recently emerged and are defeating the current Artificial Intelligence technologies. This brief analysis focuses on providing an overview of some programs aimed at protecting smart grid applications, as well as some of the challenges they face, including possible future directions for Artificial Intelligence research.