ChatGPT Creator will launch the <x0 most powerful model so far” after delay

According to reports, the White House has given the green light for publication following technical tests and discussions with OpenAI directors. The company describes GPT5.6 as its most advanced model so far, with greater skills in areas like cyber security, biology and autonomous tasks of artificial intelligence.

The delay came after US authorities worried that the powerful Ali models could be misused, for example, to discover weaknesses in software that can be exploited by foreign actors, including Russia and China.

The OpenAI has declared that it is co-operating with the US government to create a security process before publishing new models, but has warned that such controls should not become a long-term standard, as they may limit the use of technology by citizens and businesses.

GPT-5.6 will come in three versions: Sol ) the company's most powerful model, focused on maximum performance, Terraa average version of daily use, Luna more economical version, designed for speed and efficiency.

OpenAI says the new model will also have an advanced defence system with several security levels to prevent its use of cyber attacks or harmful activities, writes DW.