US government reviewing frontier AI models before release

The US government is moving to treat frontier AI models like advanced semiconductors, requiring review before release. This regulatory shift directly impacts enterprise builders using models like Claude, GPT, and Gemini. The controls aim to prevent adversarial access via open-weight releases while allowing API access with guardrails.