A new phishing attack uses hidden AI prompts in the code of emails to deceive automated security systems.

The phishing attack analyzed by security researcher Anurag revealed a new insidious tactic. The phishing message contained hidden AI prompts in the code, intended to mislead automated defense systems. Instead of standard methods, such as emergency messages or false login notifications, the attackers attempted to confuse the defense systems. The text in the email mimicked a warning that the Gmail password was about to expire, prompting users to 'confirm' their login credentials.