is now LinkedIn Learning!

To access courses again, please join LinkedIn Learning

All the same content you know and love

Plus, personalized course recommendations tailored just for you

All the same access to your Lynda learning history and certifications

Try LinkedIn Learning for free
Questions? Visit our help center.
Skip navigation

Vulnerability to attacks

Vulnerability to attacks: AI Accountability Essential Training
Vulnerability to attacks: AI Accountability Essential Training

AI algorithms can be hacked by malicious actors. For example, carefully chosen visual noise can lead an algorithm to misidentify images of dogs as images of toasters, or images of randomly chosen people as suspected terrorists. Malicious actors can also feed sufficiently large quantities of deliberately mislabeled data to render an adversary's AI algorithm useless.

Resume Transcript Auto-Scroll
Skill Level Intermediate
2h 21m
Show More Show Less
Skills covered in this course
Big Data IT

Continue Assessment

You started this assessment previously and didn't complete it. You can pick up where you left off, or start over.

Start My Free Month

Start your free month on LinkedIn Learning, which now features 100% of courses. Develop in-demand skills with access to thousands of expert-led courses on business, tech and creative topics.

Start My Free Month

You are now leaving and will be automatically redirected to LinkedIn Learning to access your learning content.