It’s a blueprint for regulatory capture, something both OpenAI and Anthropic need, in order to stay relevant. The actual doc linked in there talks about increasing the power of the Center for AI Standards and Innovation (CAISI), and having mandatory evaluations for “frontier” technology. The doc makes no mention of open source AI, privacy, freedom, or things like that.

Anthropic and OpenAI have no moats. All they can do is use fear and control to create moats.

The less OpenAI says out loud the better for everyone including themselves.