Meta and Anthropic are set to review a voluntary AI security framework with the US government, aiminβ¦