OpenAI prepares to launch its most powerful GPT-5.6 model

OpenAI, the creator of ChatGPT, is preparing to launch its new GPT-5.6 model on July 9. The previously scheduled launch was delayed in June due to additional security risk assessments by U.S. officials, as reported by Deutsche Welle.

It is reported that technical testing has concluded, and following negotiations between company leadership and U.S. officials, the White House has authorized the model's release. OpenAI describes the new GPT-5.6 as the most powerful AI system in the company's history. The model is noted for its significantly enhanced capabilities in cybersecurity, biology, and autonomous task execution.

Why was the launch delayed?

According to sources, the delay was caused by growing concerns in Washington regarding the potential for advanced AI systems to be used for illegal or dangerous purposes.

In June, the White House ordered OpenAI's competitor, Anthropic, to temporarily halt the use of its most advanced Fable and Mythos models due to national security concerns. These restrictions were reportedly eased only last week.

U.S. officials believe that modern AI models have the capability to identify vulnerabilities in software. There is concern that foreign militaries or intelligence services, including those of countries like Russia and China, could exploit this for malicious purposes.

Consequently, the White House requested that OpenAI temporarily delay the launch, a request the company agreed to. However, in its statement, OpenAI noted that such government-led reviews should not become standard practice.

Washington strengthens security measures

At the beginning of his second term, U.S. President Donald Trump was in favor of reducing regulations in the AI sector. He believed that excessive requirements could weaken the U.S.'s ability to compete with China in technology.

However, in June, Trump signed a new executive order aimed at mitigating cybersecurity risks associated with advanced AI models. Under the document, developers are to submit their models to the government for review one month before public release. It is noted that this process is voluntary.

OpenAI stated that it is working with government agencies to develop a permanent security mechanism for future model launches.

GPT-5.6 to be released in three versions

According to the company, the GPT-5.6 model launching on July 9 will come in three versions. Sol is described as the most powerful model in OpenAI's history. Terra is a mid-range variant designed for daily tasks, while Luna is the most affordable version, focusing on high speed and efficiency.

It was also announced that GPT-5.6 features a multi-layered security system. The company emphasizes that these protective mechanisms were designed to prevent the use of AI in cyberattacks or other malicious activities.

Previously, it was reported that Malta would provide ChatGPT Plus to its citizens for free, becoming the first country to offer such an opportunity. Under the agreement signed with OpenAI, Maltese citizens and residents will have free access to ChatGPT Plus for one year, requiring registration through the country's online identification system.