Scammers Use Hidden Text to Bypass AI Email Filters in Phishing Scams

Sublime Security details a rising email-based evasion technique—indirect prompt injection—where attackers hide benign text (zero-font or color-matched) in phishing emails to mislead AI/ML email filters; researchers observed two in-the-wild campaigns (an Adidas newsletter clone and a healthcare scam) and warn this nuanced attack could grow as agentic mailboxes become common.