US regulators force AI labs into limited preview, threatening growth
The U.S. government is now reviewing frontier models like Anthropic’s Mythos and OpenAI’s upcoming GPT‑5.6 before they can go public, throttling releases to a customer‑by‑customer preview. That bottleneck could stall revenue, delay data‑center build‑outs, and push the whole industry toward collective safety governance.