I did read the article. Several of the organizations mentioned simply don't talk about openness, and are instead talking about any model with sufficiently powerful capabilities, so it's not obvious why the article is making their comments about open models rather than about any model. Some of the others have made side comments about openness making it harder to take back capabilities once released, but as far as I can tell, even those organizations are still primarily concerned with capabilities, and would be comparably concerned by a proprietary model with those capabilities.

Some folks may well have co-opted the term "AI safety" to mean something other than safety, but the point of AI safety is to set an upper bound on capabilities and push for alignment, and that's true whether a model is open or closed.