From the course: Security Risks in AI and Machine Learning: Categorizing Attacks and Failure Modes
Join today to access over 24,800 courses taught by industry experts.
Attacks vs. unintentional failure modes
From the course: Security Risks in AI and Machine Learning: Categorizing Attacks and Failure Modes
Attacks vs. unintentional failure modes
- [Instructor] If one of the apps on your phone crashes, do you immediately worry, oh no, is my phone under attack by cyber-criminals? Or do you maybe shake it off and think the developers didn't test it for stability or maybe even that your phone is due for a reboot? The reality is that apps and systems can fail for both intentional and unintentional reasons. A group of experts at Microsoft and Harvard University published a taxonomy of failure modes in machine learning that describes both types of failures. And I use their research frequently as a reference point when preparing this class. Although both intentional and unintentional attacks lead to system malfunction, the underlying root cause may require a different mitigation approach. Intentional attacks are adversarial. In other words, someone is on purpose trying to disrupt the operation of the system or to use the system to their own advantage. A common adversarial…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.