Perhaps Anthropic was right in needing more oversight in using AI for murder

Maybe. That statement is doing a lot of heavy lifting. It’s a different argument than saying Anthropic’s restriction would have prevented it (it was based on bad intel/old military installation)