Abstract

Our society has transitioned from our primitive lifestyle to soon, an increasingly automatic one. That idea is further exemplified as we shift into an AI era, better known as Artificial intelligence. Artificial Intelligence is classified as computer systems that can perform tasks that typically require human intelligence. However, a common thought or question that most might have is, how is this done? How does AI process information the way we want it to and have access to so much information? AI is trained by systems called AI models. These modeling programs are trained on data to recognize patterns or make decisions without further human intervention. One AI model I will discuss in more detail is Deep Neural Networks. Deep learning is a type of machine learning that uses complex networks called deep neural networks to imitate how our brains make decisions, and though this is helpful for workflow and personal tasks, this can also become a security issue. How? Being that AI models are systems that train AI, if the AI model is compromised, that could affect the way the AI runs and functions. Though several methods can negatively affect AI models, I focus on three attacks. They are Adversarial attacks, Deserialization attacks, and Misconfiguration attacks.

Faculty Advisor/Mentor

Charlie E. Kirkpatrick

Document Type

Paper

Disciplines

Artificial Intelligence and Robotics | Educational Assessment, Evaluation, and Research

DOI

10.25776/ghpw-t252

Publication Date

4-12-2024

Upload File

wf_yes

Download

Share

COinS
 

The Security of Deep Neural Networks

Our society has transitioned from our primitive lifestyle to soon, an increasingly automatic one. That idea is further exemplified as we shift into an AI era, better known as Artificial intelligence. Artificial Intelligence is classified as computer systems that can perform tasks that typically require human intelligence. However, a common thought or question that most might have is, how is this done? How does AI process information the way we want it to and have access to so much information? AI is trained by systems called AI models. These modeling programs are trained on data to recognize patterns or make decisions without further human intervention. One AI model I will discuss in more detail is Deep Neural Networks. Deep learning is a type of machine learning that uses complex networks called deep neural networks to imitate how our brains make decisions, and though this is helpful for workflow and personal tasks, this can also become a security issue. How? Being that AI models are systems that train AI, if the AI model is compromised, that could affect the way the AI runs and functions. Though several methods can negatively affect AI models, I focus on three attacks. They are Adversarial attacks, Deserialization attacks, and Misconfiguration attacks.