Articles
Evaluating Security Risk in DeepSeek and Other Frontier Reasoning Models
5 min read
The performance of DeepSeek models has made a clear impact, but are these models safe and secure? We use algorithmic AI vulnerability testing to find out.
Using AI to Automatically Jailbreak GPT-4 and Other LLMs in Under a Minute
4 min read
The automated Tree of Attacks with pruning (TAP) method can jailbreak advanced language models like GPT-4 and Llama-2 in minutes, so they make harmful content.
28