Spammers Hijack AI Evasion Tactics to Bypass Email Filters
Spammers have adopted 'ASCII smuggling,' a technique originally designed to hide malicious prompts from AI models, to evade email security filters. By inserting invisible Unicode characters into financial keywords like 'funding,' attackers disrupt machine learning classifiers while keeping the text visible to human recipients. Microsoft detected a massive spike in this activity in early 2026, with daily detections reaching 2.5 million messages before declining in May.