"Cybertelligence": How AI is changing the face of cybersecurity
summary
Overview of concrete claims and tools: maps AI onto NIST and the OODA loop, uses supervised models for malware, spam and IDS and unsupervised for DNS analytics, and cites 70% supervised learning adoption; gives a study that AI pipelines reduced 600M raw events to 100 incidents for tier‑one analysts; flags deep learning needs massive labeled data, dark web scraping finds ~60% illicit content and stolen data up 20% since 2016, mentions haveibeenpwned and a 500,000‑device password leak and shows an adversarial image flip (panda→gibbon) as attack evidence.
AI-written summary · about
chapters (25)
- 0:00Intro
- 0:16About Me
- 1:30NIST Cybersecurity Framework
- 3:10The OODA Loop
- 5:49Applied ML for Cybersecurity
- 8:29Aspects for ML in Cybersecurity
- 9:43Neural Networks
- 10:22Bayesian Networks (How they work)
- 11:34How Does ML Work
- 12:22Major Machine Learning Uses
- 13:11Deep Learning in Cybersecurity
- 13:37Al Advantage in Cybersecurity
- 13:59What's on the Dark Web?
- 14:43Al Applications on Fighting Dark Web
- 15:24Al Fights Cyber Terrorism
- 16:37Examples
- 17:00Common Data Science Ethical issues
- 18:15Causes of Al Bias
- 19:11How Does That Happen
- 20:19Bias-variance trade-off
- 21:56Fears of Al Bias in Cybersecurity
- 22:40What's and Adversarial Attack?
- 23:45Humans
- 24:19Machines
- 24:59A Better World
related talks
11mAI-Powered Cybersecurity in the Age of Advanced Threats
44mAI Meets Cybersecurity: Boosting Speed Without Losing Control
16mIntelligent by design: Can cybersecurity cope with AI?
32mAI/Machine Learning: What is Actually Working in Cybersecurity?
49mThe Intersection of AI and Cybersecurity
01h 24mOffensive Cybersecurity by AI Promises and Pitfalls
49mGenAI and Cybersecurity: Safeguarding the Digital Age
53mHow AI is impacting cybersecurity. Does it?