Governments examined how to balance AI innovation with national security after restrictions on Anthropic’s models highlighted geopolitical fault lines.