Space Situational Awareness (SSA) plays a critical role in monitoring space debris and ensuring safe satellite operations. This dissertation introduces TLE-SafeguardNet, a novel framework designed to protect the privacy of Two-Line Element (TLE) data, a key component for accurate SSA. The proposed method combines Singular Value Decomposition (SVD) with a forward diffusion noising process, which enhances the confidentiality of TLE data while preserving its utility for downstream tasks such as satellite classification. To evaluate the effectiveness of this framework, we train a multilayer perceptron (MLP) network on the 2023 Q4 TLE dataset. Visualization techniques like t-distributed Stochastic Neighbor Embedding (t-SNE) reveal that the MLP successfully clusters payloads and debris based on eight dominant features extracted from the data. Moreover, SHapley Additive exPlanations (SHAP) plots provide interpretability by highlighting the most important features influencing the model’s decisions. The robustness of the method against noise is evaluated using a random forest classifier with 10-fold cross-validation, showing that noise can be introduced up to 401 time steps without significantly impacting data integrity. While SSA data is crucial for satellite operators to avoid collisions, it is also vulnerable to exploitation by adversaries who can use it to launch both kinetic and non-kinetic attacks on satellites. Traditional methods, such as labeling a satellite as debris, aim to hide its true identity. However, adversaries can often deduce the satellite’s actual label based on orbit information. In furthering this work, we also propose a novel adversarial machine learning approach that conceals the true identity of sensitive satellites in SSA data by modifying both the satellite’s label and orbit information. Our experimental results show that this method reduces an adversary’s ability to correctly identify a concealed satellite’s true label to less than 50%, thus improving space data security and safeguarding satellite operations from potential adversarial threats.