@rolf@drostan.org

I never thought about it like that: current phishing emails are usually very obvious, because a scammer wants to limit the replies to only include the most promising leads. The subsequent exchange of messages is labour-intensive. But Large Language Models make it possible to scale this, even using available information about a victim.
#LLMs #ChatGPT

"LLMs and Phishing"
https://www.schneier.com/blog/archives/2023/04/llms-and-phishing.html