This makes sense if Anthropic think they're the best-positioned to make safe AI. However if you are looking at an AI company there's obviously some selection happening.