A study provides a new way of prompt injection into LLMs. It uses a structural issue and therefor is highly successful on jailbreak benchmarks (around 60%).
All you have to do, is hijack the LLMs 'reasoning' stage by forming your text as if it was a conclusion by the LLM itself, which leads to the LLM to trust this injection as if it was an own conclusion.
In march meta introduced a new ai chatbot on instagram that should help users to solve problems with their account.
And so it did. Problem is: it helped any user that asked nicely, to recover any account.
There once was a time, when adding 'crypto' to your company name, boosted your share values. Now is the time, when adding 'AI' to your company name is boosting your share value.
And now finally the two awesomest shit-technologies of the 21st century learned to work together and helped to steal 150000 USD worth of cypto tokens from a wallet.
As reported by Toms Hardware the owner of PocketOS, a car rental SaaS platform, used a cursor ai agent to debug an authentication problem.