OpenAI Limits Release Of Sol, Terra And Luna Models Amid US Government Safety Review

OpenAI has officially introduced three new artificial intelligence models—Sol, Terra, and Luna. However, the company has deferred a broad public rollout, opting instead for a restricted preview among a select group of trusted partners, following a specific request from the US government for an early, comprehensive safety review of these frontier systems.

This move marks a significant moment in the intersection of AI development and national policy. By seeking early access to these frontier models, the US government is signaling a more proactive stance on the potential risks associated with advanced AI architectures. The decision to delay the public release highlights the ongoing efforts to balance rapid innovation with controlled deployment protocols.

The restricted launch reflects a broader industry trend where major AI labs are increasingly coordinating with policymakers before releasing highly capable systems. The outcome of the ongoing safety review by government agencies will likely dictate the timeline for when these models become accessible to a wider user base, setting a precedent for how future generation frontier models are governed and deployed.