Here’s how attackers are getting around phishing defenses

Egress intelligence shows attackers are evading NLP-based email security by appending benign text, legitimate-looking links (e.g., Bank of America, Uber), random characters, and whitespace to malicious emails so automated processors classify them as safe; analysis of 40 US-targeted attacks found frequent use of multiple obfuscation techniques, and the report notes scanning timeouts and stacking of benign elements enable delivery, with broader data indicating phishing remains a substantial portion of incidents.