AI-Powered Solutions for Cybersecurity
summary
Claims a production phishing detector that holds 95% average accuracy since March 2022, trained on combined sources including PhishTank and OpenPhish to a six‑digit dataset. Core pipeline tokenizes and vectorizes URLs into a neural net, falls back to on‑page scraping when needed using aiohttp asynchronous requests plus BeautifulSoup with lxml and cchardet for faster parsing. Also reports NLP plus computer‑vision steps for display‑name/lookalike spoofing and a live URL checker.
AI-written summary · about
chapters (14)
- 0:00AI in cybersecurity intro
- 1:36AI learning paradigms
- 3:46Spoofing attack detection
- 7:23Phishing website threats
- 9:31Phishing data specifics
- 12:43URL data preprocessing
- 14:57URL feature extraction
- 18:12Advanced phishing detection
- 20:00Async URL requests
- 23:18Web scraping optimization
- 24:21Model evaluation metrics
- 25:44Deployment results
- 26:25Future AI challenges
- 27:07Q&A
related talks
51mDeepPhish: Simulating Malicious AI
11mAI-Powered Cybersecurity in the Age of Advanced Threats
51mDeepPhish: Simulating Malicious AI
21mMachine Learning and Cyber Security - Detecting malicious URLs in the haystack
30mBuilding a Predictive Pipeline to Rapidly Detect Phishing Domains
11mHow to Navigate Emerging AI-Driven Security
50mCatPhish Automation - The Emerging Use of Artificial Intelligence in Social Engineering
09mBridging the Gap: Developing Accessible Anti-Phishing Solutions