ai needs a system like white/grey hat and this 'exclusive' red team stuff is not it. eg what if, i am a 18yo kid who can jailbreak llms for fun. the current system will not let me in b/c I am not an 'ai researcher' and most prolly my attempts will flag my account and get me in trouble with authorities.

ref: https://www.anthropic.com/news/model-safety-bug-bounty

If you are some rando who does this for fun I’m sure you can figure out a way to get in touch with these people

It says OR not AND so I don’t even think they are mandating formal experience

Anthropic market themselves as the more ethical version of OpenAI. But a simple scan of their respective privacy policies will show you pretty much the opposite. And your screenshot hints again at dubious ethics.

AI needs more open-source.. !