China and Russia now have it.

Hegseth declared Anthropic to be a national security threat. How can we use their products now? Or did the Divine Don say it was okay so all is forgiven?

I am very confused. Thank you for your attention to this matter.

It's no major leap forward in LLM tech. One review I read was very detailed, and absolutely eviscerated the source document.

1. Out of something like ~244 pages in the source doc, only 7 were dedicated to the security vulnerability assessment. There was no mention of the much-touted "thousands" of zero-days claimed in the media. None.

2. Those 7 pages of security assessment were in turn based on something like ~54 already known vulnerabilities in Firefox version whatever that were identified by Anthropic Opus model. They basically took Opus output and gave it to Mythos to work on. 54 NOT "thousands".

3. Mythos did not assess Firefox version X, at all. It assessed a un-sandboxed javascript runtime environment that emulated an unsecured Firefox version X.

4. By the time Anthropic ran this fake test, Firefox had already released a security update plugging the 54 holes that Opus (not Mythos) had previously identified.

To sum:

No "thousands".
Not "Mythos".
Not "Firefox".
Not "zero-day".

Now we see Anthropic getting a gov't contract, after all, and Mythos is being hidden from the public and touted as a huge step forward.

This is how Anthropic is getting around the pipeline ban that Kegsbreath put in place after they wouldn't cow to his demand for robot soldiers.