OpenAI prepares to launch model after Trump administration release

OpenAI will publicly launch GPT-5.6, its most advanced artificial intelligence model, this Thursday (9), following a delay last month prompted by requests from the United States government due to increasing national security concerns regarding the possible misuse of powerful AI technologies.

The United States and China are competing for leadership in the development of cutting-edge AI models that experts say can significantly accelerate sophisticated cyberattacks against industries that rely on complex, interconnected, and often decades-old technological systems.

Washington has stepped up analysis of the launches of advanced AI models to identify potential threats, amid concerns that the technology could be misused by the armed forces or intelligence services of China, Russia and other countries.

08/07/2026

Emmy 2026: See list of nominees

08/07/2026

Apple closes R$155 billion deal to manufacture chips in the USA

08/07/2026

Wreckage of missing Boeing 737 in Pakistan is found; Search for crew continues

---

CANAL NO WHATSAPP

Chinese authorities have also held meetings with major technology companies to discuss the possibility of restricting international access to the country's most advanced AI models, including those that have not yet been launched.

OpenAI competitor Anthropic suddenly disabled its most advanced AI models, Mythos 5 and Fable 5, for all users following the export control order issued by the United States government on June 12 over national security concerns. Restrictions were lifted last week after Anthropic implemented certain protective measures.

Axios, which broke the news about OpenAI's launch, reported that the Trump administration approved a broad rollout of GPT-5.6 after additional testing and meetings between the company and government officials.

Top Cifras

Toque agora.

1 Un Trocito De Cielo Jaci Velasquez

2 Maybe someday The Cure

3 Vem Com Teu Fogo Adorazione

4 Nunca Había Sentido Tanto Miedo Alejandro Neri

The White House and the U.S. Commerce Department did not respond to a Reuters request for comment outside business hours.

OpenAI had limited access to GPT-5.6 to a small group of previously vetted partners, whose data was shared with authorities.

The company responsible for ChatGPT will launch its most advanced model, the GPT-5.6 Sol, along with the lower-cost Terra and Luna models, it said in a post on the social network X at the end of Tuesday.

OpenAI highlighted improvements in the capabilities of AI agents aimed at programming, biology and cybersecurity when it previewed the models in late June. At the time, the company stated that GPT-5.6 Sol was competitive with Anthropic's Mythos Preview in the ExploitBench cybersecurity benchmark.

Billionaire Elon Musk, whose company SpaceXAI (SPCX.O) competes with Anthropic and OpenAI, said on Wednesday that his company would also make its flagship model, the Grok 4.5, available to the public.

National security concerns
US President Donald Trump signed an executive order establishing a voluntary framework by which AI developers can provide "covered frontier models" to the US government for up to 30 days before making them available to partners deemed trustworthy.

Although Washington has lifted export controls on Anthropic's Fable model, Mythos, designed for cybersecurity professionals, remains available only to a few U.S. organizations deemed "trusted."

In China, authorities are concerned about Mythos' potential to exploit software vulnerabilities and the possibility of the United States using the model against Beijing's interests.

Anthropic warned that it is "probably impossible" to make any artificial intelligence model fully resistant to jailbreak techniques.

See the states that most use ChatGPT in Brazil, according to OpenAI

Source: CNN