I don't think there's a government in the world, including the US, that should allow Palantir anywhere near their data or systems. I consider Palantir a national security threat. I also feel this way about McKinsey (and Bain, BCG, etc).

I also think any form of platform AI usage to be a national security threat in the absence of stringent controls over that data and the platform. At some point I think governments and companies will wake up to this and demand local LLMs or, in the very least, a cloud platform within their jurisdiction, ownership and control.

The 1980s and 1990s ushered in this idea of "small government", privatization and public-private partnerships that I think was a huge mistake with catastrophic consequences. It's simply letting the foxes into the hen house. It leads to regulatory capture, a revolving door and a massive government-to-private wealth transfer.

What's funny is that a lot of this stems from a now throughly debunked idea of the "tragedy of the commons" [1].

[1]: https://www.usnews.com/opinion/economic-intelligence/2015/03...