OpenAI Has New AI Models. Here’s Why You Can’t Use Them

2026-06-26 · Source: WIRED - Ai · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Emerging Technologies & Innovation · Depth: Fundamental Awareness, short

Summary

OpenAI has delayed the public release of its new GPT-5.6 AI models, including Sol, Terra, and Luna, following a request from the White House. This delay, which OpenAI views as temporary, comes two weeks after Anthropic also had to take its advanced AI models offline due to government intervention. The White House's actions are driven by increasing concerns over the cybersecurity capabilities of powerful new AI models, despite a recent executive order aiming for a "voluntary process" for AI labs to share models 30 days pre-release. OpenAI states this voluntary framework is not yet established, creating an uncertain interim period for US AI labs. GPT-5.6 Sol is described as OpenAI's most capable model on benchmarks for cybersecurity, biology, and agentic abilities, and will feature a "layered safeguard stack" to prevent malicious use. OpenAI plans to broaden customer access, including international partners, in the coming weeks.

Key takeaway

For Directors of AI/ML planning new model releases, you must now factor in potential government pre-release review, even if a formal "voluntary process" is undefined. Your release timelines for advanced AI models, especially those with cybersecurity implications, could face delays similar to OpenAI's GPT-5.6 or Anthropic's models. Proactively engage with regulatory bodies and develop robust internal safeguard stacks to mitigate risks and navigate this evolving, uncertain regulatory landscape.

Key insights

The US government is imposing de facto pre-release review on frontier AI models due to cybersecurity concerns.

Principles

• Government intervention can delay frontier AI model releases.
• Cybersecurity risks are driving new AI model regulations.
• Voluntary AI model review processes may become mandatory.

Method

The article describes a government review process where AI labs submit model lists for approval before broader release, but no formal voluntary framework exists yet.

In practice

• Develop "layered safeguard stacks" for AI model security.
• Prepare for government pre-release model reviews.

Topics

• AI Model Regulation
• OpenAI GPT-5.6
• Cybersecurity Concerns
• White House Executive Order
• Frontier AI Models
• Anthropic

Best for: CTO, VP of Engineering/Data, Executive, Policy Maker, Director of AI/ML, Tech Journalist

Related on AIssential

• OpenAI limits GPT-5.6 rollout after government request, says restrictions shouldn’t be the norm — Government intervention is delaying advanced AI model releases, prompting industry pushback and calls for clear regulatory frameworks.
• OpenAI staggers AI model release after Trump administration request — Government oversight is intensifying for advanced AI model releases due to national security and capability concerns.
• Trump's new executive order wants AI companies to voluntarily submit models for government safety reviews — Government seeks voluntary AI model safety reviews while major labs advocate for mandatory national frameworks.
• US reportedly urges Meta to submit AI models — US government seeks pre-release AI model evaluations from major developers for national security and safety.
• What CIOs should watch for in Trump’s AI oversight order — Voluntary federal review of frontier AI models prioritizes national security and cybersecurity without mandatory licensing.
• White House reins in OpenAI's GPT-5.6 — Government oversight is emerging as a critical gatekeeper for releasing powerful, "Mythos-level" AI models to the public.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by WIRED - Ai.