callumzc

Glasswing was an overlooked governance precedent. Through it, Anthropic recognised that when capabilities rapidly advance and frontier models could cause serious harm, they have a responsibility to control who gets to access their tech. Here, I deal with the governance implications from Anthropic taking on that responsibility. (I also note...

Last month, Anthropic developed Claude Mythos, a model they considered too dangerous for public release. As per Anthropic (and via testing from AISI), we know that Mythos: * Found thousands of previously unknown vulnerabilities in every major operating system and browser. * Surpasses the coding capabilities of all but the...