The majority of attacks against LLMs come from text based prompts that will cause the agent to ‘jump the rails’ and start providing responses it shouldn’t.  An older example of this was…

Source: Read More