It Is The Year 200, We Are Robots
summary
Claim: offensive machine learning automates and scales phishing ops using real tools and measurable rates. Projects include op bot, an RL tool for privilege hunting, proof pudding, an adversarial model that targets Proofpoint, and deep drop refactored into a sandbox-scoring API. Uses GPT-style generators to create varied emails (about 30% usable after edits). Operational stats cited: 10% sandboxing, ~80% delivery, ~60% click rate; operators manage 4–6 personas and often use generated faces from thispersondoesnotexist.
AI-written summary · about
chapters (9)
- 0:00Introduction and talk overview
- 2:10Machine learning fundamentals
- 4:59Offensive machine learning defined
- 11:01Phishing challenges today
- 14:29Phishing persona strategies
- 19:31Target selection and engagement
- 22:42AI for phishing email generation
- 25:31AI for chatbot phishing
- 28:50Q&A session
