"AI Under Criminal Influence: Adversarial Machine Learning Explained"

Since the release of ChatGPT to the public, the adoption of Artificial Intelligence (AI) and Machine learning (ML) systems has increased significantly. In order to gain a competitive advantage, companies are racing to adopt AI technology. However, they may be exposing themselves to cybercriminals. ML models that drive many AI applications are susceptible to attacks against data contained within AI systems and adversarial ML attacks. Adversarial ML involves providing malicious input to an ML model to cause it to generate inaccurate results or degrade its performance. This attack could occur during the training phase of the ML model, or it could be introduced later via input samples to trick a trained model. This article continues to discuss how ML models are trained, the different types of adversarial ML attacks, and how to combat such attacks.

Cybernews reports "AI Under Criminal Influence: Adversarial Machine Learning Explained"