Paper by Anthropic outlines how LLMs can be forced to generate responses to potentially harmful requests
The safety features on some of the most powerful AI tools that stop them being used for cybercrime or terrorism can be bypassed simply by flooding them with examples of wrongdoing, research has shown.
In a paper from the AI lab Anthropic, which produces the large language model (LLM) behind the ChatGPT rival Claude, researchers described an attack they called “many-shot jailbreaking”. The attack was as simple as it was effective.
About Author
More Stories
The 10 rules of friendship: show up, go beyond banter, learn the boring details
Bizarre Australian mole even more unusual than first thought, new research reveals
Male mosquitoes to be genetically engineered to poison females with semen in Australian research