OpenAI Finally Rolls Out Its Most Advanced GPT Model After Government Delays

OpenAI is gearing up to release its most advanced language model yet, GPT-5.6, after a hold-up caused by increasing scrutiny from U.S. authorities. The launch, initially postponed last month, now has the green light amid worries that sophisticated AI systems could pose national security risks.

This move comes at a time when the U.S. and China are locked in a tight race to develop leading-edge AI technologies. Experts warn these new tools might accelerate cyberattacks, especially targeting industries with intricate or aging infrastructure.

The regulators in Washington have stepped up their oversight of AI deployments, aiming to prevent the technology from being exploited by adversarial states or malicious actors. Concurrently, Chinese regulators are contemplating limiting access to their most powerful AI models, highlighting how this battle for AI supremacy is playing out on a geopolitical scale.

OpenAI's competitor, Anthropic, recently paused access to some of its top AI offerings following U.S. government restrictions issued in June. After implementing security measures, these controls were eased but not entirely lifted, leaving some advanced tools available only to a select group of trusted organizations.

Reports suggest that the previous U.S. administration gave approval for the broad launch of OpenAI's GPT-5.6 following further company-government consultations. OpenAI had previously limited the model's use to vetted partners whose details were shared with U.S. officials to mitigate risks.

Alongside the GPT-5.6 release, OpenAI plans to introduce more affordable versions-Terra and Luna-expanding access options. Meanwhile, SpaceXAI's Elon Musk is reportedly making his Grok 4.5 model accessible to the public, signaling intensified competition in the AI market.

Security remains a heated topic. U.S. President Donald Trump recently signed an executive order encouraging AI developers to provide new frontier models to the government temporarily before wider release. The government is keen on vetting these technologies to forestall misuse in cyber warfare and intelligence operations.

While some AI tools have had their export restrictions eased, models designed for cybersecurity tasks still face limited distribution due to fears over their potential exploitation. This underscores the challenges in balancing innovation and security.

Japanese cybersecurity experts argue that making AI models entirely jailbreak-proof might be out of reach, adding another layer of complexity to their governance.