From the course: CompTIA SecAI+ (CY0-001) Cert Prep

Unlock this course with a free trial

Join today to access over 25,600 courses taught by industry experts.

Model inversion

Model inversion

From the course: CompTIA SecAI+ (CY0-001) Cert Prep

Start my 1-month free trial Buy for my team

Model inversion

“

Model inversion attacks try to use prompts to extract information about the data a model was trained on. Sometimes the goal is to recover actual examples from the training set. In other cases, the goal is to infer sensitive attributes or reconstruct inputs based on the model's responses. The model itself becomes the attack surface, even when the attacker cannot access the underlying data directly. These attacks take advantage of the way that models memorize information. If training data contains personal, medical, or proprietary content, an attacker can sometimes coax the model to repeating pieces of that training content. The model may not intend to memorize this sensitive information, but large-scale training on data that was not cleaned or filtered can leave traces that leak through to answers to prompts. One model inversion attack technique uses repeated queries with small variations in each query. The attacker keeps prompting and looks for consistent patterns that reveal hidden…

Contents