OpenAI has limited the rollout of its latest AI model, GPT-5.6, to a select group of approved partners, citing ongoing government pressure and cybersecurity reviews. This move follows calls for stricter oversight of advanced AI systems.