Claude Mythos 5.2 is in AWS GovCloud for approved US customers
It is the first Anthropic model with strong cyber functions in AWS GovCloud (US). Only approved organizations can use it.
Amazon Bedrock, a platform for AI tools, offers Claude Mythos 5.2 to eligible customers in AWS GovCloud (US). This is the Amazon cloud for US government work. Anthropic names it its most advanced model for cybersecurity, code and hard reasoning. It is the first Anthropic model with strong cyber functions in this cloud. Access is limited to eligible customers.
What it does
Amazon tells that the model finds important security defects in software and shows that an attack can use them. It reads large codebases and gives results with less manual guidance than earlier models. Security teams can use it for defence.
The proof
The text gives no test scores, no numbers and no price. The claims are from the own description of the model by Amazon.
What is not known
Access is for organizations on an approved list only. To start, an organization applies to the Cyber Verification Program of Anthropic. The text does not tell who is approved or how long it takes.
Insight: Anthropic decides who gets its top cyber model
Anthropic gave Mythos 5.1 only to verified researchers, and on 26 October it applied for a Nasdaq listing. Mythos 5.2 is in the US government cloud, but only for organizations on the approved list. The verification list is the gate, so Anthropic decides who gets the model. For security leaders, a name on that list is a planning item.
Sources
Posted
