Why We Don't Use Machine Learning for Bot Detection
An honest argument against ML-driven bot detection. Adversarial drift, training data poisoning, latency, and the explainability tax that nobody talks about.
bot-detectionInsights and updates on AI bot detection, web security, and digital threat mitigation. Discover best practices for protecting your online properties from automated threats and malicious bots.
Post 13 through 18 of 57 total posts
An honest argument against ML-driven bot detection. Adversarial drift, training data poisoning, latency, and the explainability tax that nobody talks about.
bot-detectionA technical deep dive into Anubis, the proof-of-work AI scraper blocker now running in front of GNOME, KDE, FFmpeg, and dozens of public infrastructure sites.
bot-detectionLegacy CSS honeypots like display:none and offscreen positioning no longer fool vision-based AI agents. New placement strategies for 2026.
bot-detectionAn honest, technical look at why CAPTCHAs no longer stop bots, and the modern stack of behavioral biometrics, proof-of-work, and fingerprinting that does.
bot-detectionAI-generated form spam is harder to catch than the old kind. An honest technical breakdown of what works, what fails, and where the arms race is going.
securityA technical deep dive into credential stuffing tooling, attack anatomy, and the detection signals that actually work against modern ATO operators.
security