Anthropic surely is well aware. Most likely they asked separate agents multiple times to code review the proof and look for exploits.
Anthropic surely is well aware. Most likely they asked separate agents multiple times to code review the proof and look for exploits.