From the course: Complete Guide to Cybersecurity: A Practical Approach

Unlock this course with a free trial

Join today to access over 24,400 courses taught by industry experts.

Understanding model inversion attacks

Understanding model inversion attacks

From the course: Complete Guide to Cybersecurity: A Practical Approach

Start my 1-month free trial Buy for my team

Understanding model inversion attacks

“

- In a model inversion attack, the attacker tries to infer sensitive information about the training dataset. And usually this takes form of a prediction, API, and some accelerated information that may gather from the system. Now, machine learning and AI models, in the process of learning, they try to make accurate predictions, right? In this case, they may inadvertently memorize or capture aspects of the training data. And then of course, if the data is not diverse or the model is not overfitted, the attacker can exploit this to invert the model, so that it can actually generate input that causes the model to output information that reveals something about the training data itself. So at the end of the day, what the attacker wants to do is to be able to perform some type of reconnaissance about the training data, and about the actual underlying infrastructure of the AI implementation. So let's take a look at a few scenarios in here. So let's assume that an AI model is trained to…

Contents