OpenAI Launches GPT-5.6 After Government-Requested Delay

On Thursday, July 9, 2026, OpenAI announced that its GPT-5.6 family—Sol, Terra and Luna—will be made publicly available. The company posted on X, “We’re expanding preview access globally now,” and CEO Sam Altman added “happy building.” The release follows a two-week pause that began after the Trump administration asked OpenAI to stagger the rollout because of heightened national-security concerns.

9to5Mac

In early June, President Donald Trump signed an executive order establishing a voluntary framework that asks AI developers to give the U.S. government up to 30 days of access to “covered frontier models” before broader distribution. The order was prompted by concerns that powerful models could be weaponized in cyber-attacks, a fear amplified after Anthropic’s Mythos and Fable models were temporarily blocked by the Department of Commerce. OpenAI initially limited GPT-5.6 to a “small group of trusted partners” whose participation was shared with federal officials, describing the arrangement as a “government access process” it does not view as a long-term default.

YnetNews

• Sam Altman – CEO of OpenAI, who met with House Minority Leader Hakeem Jeffries on June 3, 2026, and communicated the release on X.

CNBC

• The Trump administration – Issued the June 2 executive order and later requested the delayed rollout.

The Hill

• U.S. Department of Commerce’s Center for AI Standards and Innovation – Conducted additional testing of GPT-5.6 after OpenAI sent technical staff to Washington.

Trump administration lifts restrictions on GPT 5.6 from OpenAI

• Anthropic – Competitor whose Mythos and Fable models were subject to the same export-control scrutiny.

Gizmodo

• Model tiers: Sol (flagship, strong in coding, biology, cybersecurity), Terra (mid-range for everyday work), Luna (fast, low-cost).

OpenAI GPT-5.6 Public Rollout Starts July 9

• Pricing (reported by multiple outlets): Sol – $5 per million input tokens, $30 per million output; Terra – $2.50 input, $15 output; Luna – $1 input, $6 output.

Engadget

• Safety claims: OpenAI says GPT-5.6 includes safeguards capable of withstanding “real-world adversarial pressure.”

NextGov

OpenAI reiterated that the temporary limitation was “the strongest path to broader availability in the coming weeks” while it collaborated with the administration to develop a repeatable release framework. The company emphasized that it “doesn’t believe this kind of government access process should become the long-term default.”

NewsNation

A White House spokesperson clarified that the administration “does not provide approvals for private companies to release AI models – decisions on timing and scope of releases rest entirely with the companies,” and that the executive order “does not mandate pre-clearance.”

Gizmodo

The Department of Commerce’s Center for AI Standards and Innovation confirmed it performed additional testing but did not issue a formal approval, noting the process was voluntary.

The Hill

Policy analysts warned that the ad-hoc handling of model releases could set a “dangerous precedent” for government influence over AI innovation. Security experts expressed ongoing concern that even with safeguards, frontier models like GPT-5.6 could accelerate sophisticated cyber-attacks if exploited by hostile actors.

The Hill

• Government “green light”: Axios-cited sources reported that the Trump administration gave OpenAI permission for a broad launch after extra testing. In contrast, Gizmodo and a White House official denied any formal approval, stating no permission was required or granted.

Gizmodo

• Release date: Some outlets list July 8 as the launch date, while the majority, including OpenAI’s own X post, specify July 9.

9to5Mac

• Scope of availability: OpenAI has not clarified whether the Thursday rollout applies globally or is initially limited to U.S. users.

The Hill

• “happy building” — Sam Altman, CEO, OpenAI

The Hill

• “we don't believe this kind of government access process should become the long-term default” — OpenAI statement (June 2026)

OpenAI to release GPT-5.6 after reported US review

• “The Administration does not provide approvals for private companies to release AI models – decisions on timing and scope of releases rest entirely with the companies,” — White House spokesperson

The Hill

• “We’re expanding preview access globally now,” — OpenAI X post, July 8, 2026

NextGov

OpenAI indicated that it will continue working with the Department of Commerce to refine the voluntary testing framework outlined in the executive order, aiming to establish a repeatable process for future frontier-model releases. The company also announced the simultaneous rollout of GPT-Live voice models, which can listen and speak concurrently, signaling broader product expansion beyond the GPT-5.6 series.

CNBC