So this hack was via China's LLM apparently
From what I just heard on Natalie Brunell's podcast, this was caused by someone using China's brand new LLM that ripped off code from Anthropic. It has only been out a few days. Unlike Anthropic the China model dows not have any safeguards against people using it for bad things. The guest on the podcast stated that he was able to replicate the hacking code in 90 seconds. Seconds!! So clearly now that the word is out, many bad actors will try to replicate the LLM model to find flaws in code on every device. This is not over folks, it's just starting.
[link] [comments]