Archive — US — techcrunch
OpenAI limits GPT-5.6 rollout after government request, says restrictionsshouldn’t be the norm
OpenAI is limiting the rollout of its GPT-5.6 models, including Sol, Terra, and Luna, to a small group of trusted partners following a U.S. government request for tighter controls over advanced AI releases. This move raises concerns about government overreach in AI regulation and could impact global competitiveness in the AI race, while OpenAI expresses frustration at these restrictions, stating they should not be the norm.
Key points
- OpenAI to limit GPT-5.6 rollout to selected partners due to government request.
- Models include Sol (flagship), Terra (balanced), and Luna (lower-cost).
- The Trump administration's restrictions raise concerns over governmental power in AI.
- A former White House adviser criticizes the executive order as creating a de facto licensing regime that delays innovation.
- GPT-5.6 Sol has advanced capabilities in coding, biology, and cybersecurity.
- OpenAI emphasizes the importance of balancing model safety with accessibility.
Why it matters
- In 2023, other companies, like Anthropic, faced similar restrictions post model releases, illustrating industry-wide tensions between AI innovation and regulatory concerns.
- Discussion around government control over AI has parallels to prior debates on privacy and tech regulation, which continue to evolve leading into 2025.
While OpenAI complies with government requests for limiting GPT-5.6, the company warns this approach could hinder future advancements in AI technology, threatening US competitiveness and innovation in a rapidly evolving landscape.