Tag Archives: generative AI

Malicious use of deepfakes: Attacker can take advantage of using AI for producing a fake voice, video. Data leaks that expose confidential corporate information, extracting sensitive training data(‘model inversion’): API key, IP, source code, sensitive training data in general, PII. Data poisoning (corrupting training data): Attacker can inject a lot of fake news, fake informationContinueContinue reading “New risks from generative AI”