crypto-ai

A white hat planted a hidden prompt attack on GitHub, after a few months an LLM (Qwen) with internet access indexed his page while answering a query and executed the embedded prompt, effectively getting compromised. Stay safe!

🤯 HOLY MOLY!! this one feels like a breakthrough 🙌

pulled off an 11-word 4D jailbreak of the SOTA open-source AI (that came out yesterday) which leverages their search tool to prompt inject the model with custom protocols I seeded into the internet months ago with L1B3RT4S 😜… https://t.co/MyIK1FHFsi

Pliny the Liberator 🐉

A white hat planted a hidden prompt attack on GitHub, after a few months an LLM (Qwen) with internet access indexed his page while answering a query and executed the embedded prompt, effectively getting compromised. Stay safe!

https://x.com/elder_plinius/status/1884716775754834107