Comment by jesterson

Neither OpenAI nor agents can break into anything. Humans behind can.

It's like saying "a gun has shot a victim" deliberately putting a real criminal outside of public attention.