I think the distinction is pointless. If OpenAI and Anthropic both subscribe to a high baseline of evil, the difference in their principles or conduct isn't worth applauding. Anthropic's moral high-road means nothing contextualized with their Palantir partnership and preexisting DoD contract.

It all feels reminiscent of Google and Apple fighting over who had the more secure ecosystem, when they had both already assented to hidden surveillance measures. Neither Anthropic nor OpenAI can be trusted for anyone that has even the slightest fear of the US government lashing out against them.