LLMs have been observed leading users to phishing links. 

LLMs have been observed leading users to phishing links. 

Source: cybersecurity insiders
Source Link: https://www.securitymagazine.com/articles/101738-phishing-scams-can-deceive-large-language-models