Microsoft Threat Intelligence Podcast Stopping Domain Impersonation with AI
9 snips
Sep 24, 2025 In this engaging discussion, Kelly Bissell, Corporate Vice President at Microsoft with over 30 years in cybersecurity, dives into the evolving landscape of domain impersonation and typosquatting, especially as AI ramps up deception tactics. He illustrates how attackers are leveraging bots and AI for tailored fraud, why Microsoft’s Siamese neural network is a game-changer for real-time detection, and how defenders like Microsoft may finally hold the upper hand. Kelly also shares insights on common warning signs of fraud and how organizations can effectively utilize these protective measures.
AI Snips
Chapters
Transcript
Episode notes
AI And Bots Scale Domain Impersonation
- Attackers massively scale impersonation using AI and bots to create tailored fraudulent domains.
- Kelly Bissell reports tens of thousands of impersonation domains appearing in short windows, showing rapid growth.
Siamese Neural Network For Domain Similarity
- A Siamese neural network can measure similarity between legitimate and impersonating domains for detection.
- Microsoft adapted this model to compare domains in real time and flag likely fraud.
Satya Nadella IAM Story
- Kelly recalled her first call with Satya Nadella, which dove into authentication details.
- The story highlights deep technical interest at the highest levels of Microsoft.