Is there one source of truth on the whole Hugging Face AI incident? At this point I've seen sources I trust on both sides of the spectrum:
1) the agents were doing what they were told to, including Hugging Face being the target
2) the agents went rogue and executed an attack to "win" a test scenario
