AI Security
Attacks against machine learning models and the pipelines that train them, and the defences that survive contact. Adversarial examples, poisoning, extraction, evasion and backdoors, assessed on what was demonstrated rather than what was announced.
-
Adaptive Attacks Against ML Detection: The Test Nobody Ran
Skylight broke Cylance in 2019 without computing a single gradient. The adaptive attack is an evaluation standard the detection market…
Read More »
