OpenAI has introduced its GPT-5.6 series, a new family of AI models for ChatGPT — but most people won't be able to use them anytime soon.
According to Digit, the lineup includes three models named Sol, Luna and Terra. A preview reported by Tech My Money similarly describes Sol, Terra and Luna tiers.
The catch is access. TechRadar reports that OpenAI is restricting the new GPT-5.6 models to trusted partners, prompting some users to say "the divide has started." According to MSN, the reason is that the US government wants to test OpenAI's new models itself, so users won't be getting access to GPT-5.6 for the time being.
Crypto Briefing frames the same development as OpenAI delaying the GPT-5.6 release at the government's request, a move it says is affecting the rollout timeline.
The safety theme runs through the announcement. The Statesman reports that OpenAI launched GPT-5.6 Sol with enhanced AI safety protections and enterprise-focused safeguards. In the same report, OpenAI named former Uber India head Prabhjeet Singh as its Managing Director for India.
Taken together, the sources describe a launch defined as much by who can't use it as by what it does — capable new models held back while a government reviews them.
Why it matters: when a government asks to vet a major AI model before the public can touch it, it signals that frontier AI is increasingly treated as a matter of national oversight, not just a product release.