OpenAI Finally Gets the Green Light — GPT-5.6 Drops This Thursday

TLDR

• OpenAI will publicly launch GPT-5.6 — including Sol, Terra, and Luna models — on Thursday after US government approval
• The launch was delayed due to national security concerns over AI’s ability to identify software vulnerabilities
• The Trump administration approved the broad release following testing and meetings with OpenAI officials
• Rival Anthropic had its Mythos 5 and Fable 5 models briefly disabled after a June 12 export control order, with access partially restored last week
• Both OpenAI and Anthropic have filed confidential IPO documents targeting valuations near $1 trillion

OpenAI is set to publicly release its most powerful AI model series, GPT-5.6, on Thursday after the US government gave it the green light following national security reviews.

The GPT-5.6 lineup has three tiers. Sol is the flagship model. Terra is a mid-range option for everyday work. Luna is a fast, low-cost version.

OpenAI confirmed the launch in an X post on Tuesday, saying it was also expanding preview access globally ahead of the full release.

The launch had been delayed after the US government asked OpenAI to limit access to a small group of vetted US-based partners. That request came after concerns that advanced AI models could help hackers identify vulnerabilities in code.

Axios reported the Trump administration approved the broader launch after additional technical testing and meetings between OpenAI and government officials.

National Security Concerns Shaped the Release

Washington has been stepping up oversight of frontier AI models over fears the technology could be exploited by adversaries, including China and Russia.

The US and China are in a race to develop cutting-edge AI, and experts say these models could accelerate sophisticated cyberattacks on complex, aging infrastructure.

OpenAI rival Anthropic faced similar restrictions. Its Mythos 5 and Fable 5 models were disabled for all users after a US export control order on June 12. Access was partially restored last week after Anthropic put safeguards in place.

Mythos, which is built for cybersecurity professionals, remains available only to select trusted US organizations. China has raised concerns the model could be used to exploit software vulnerabilities against its interests.

Anthropic has said it is “probably impossible” to make any AI model fully resistant to jailbreaks.

Pricing and Competition

OpenAI said Terra will be priced at half the cost of its predecessor, GPT-5.5, as competition with Anthropic and Google intensifies.

GPT-5.6 Sol was described by OpenAI as competitive with Anthropic’s Mythos Preview on the ExploitBench cybersecurity benchmark when previewed in late June.

Elon Musk’s SpaceXAI also moved this week, making its Grok 4.5 model available to the public.

The Trump administration has signed an executive order creating a voluntary framework where AI developers can share frontier models with the US government for up to 30 days before broader release.

OpenAI has said it does not want this kind of government access process to become the long-term default, arguing it keeps powerful tools away from businesses and users who need them.

Both OpenAI and Anthropic have filed confidential IPO documents with US regulators, with both companies targeting valuations close to $1 trillion.

The post OpenAI Finally Gets the Green Light — GPT-5.6 Drops This Thursday appeared first on CoinCentral.