Kinda like ChatGPT hired a person to just pass captchas, or Sydney reading news about it's own actions/interactions and therefore getting info it shouldn't have had, there's a lot of space for going out of the guardrails of not having a proper interface

Too many systems rely on people being unaware of exploits, but an AI would never forget something, or get bored or tired of trying, It doesn't need to be smarter than humans, just have enough persistence and attention to detail