GPT-6 Cyber should be officially launched over the coming weeks and is the fourth model of this type that OpenAI offers in 2026. It is a sign that cybersecurity is an increasingly important issue for those working in Artificial Intelligence.
© Shutterstock
25/09/2026
by
A cybersecurity has been an increasingly raised issue among those who are concerned with the development of Artificial Intelligence, with the concerns to be stressed after the cases in which OpenAI, Anthropic and Google models managed to invade systems from other companies.
A Fortune it says now that OpenAI has plans to unveil GPT-6 Cyber, a new model of its cyber-security-focused Artificial Intelligence. According to publication sources, this new model will be launched in the coming weeks.
It seems that OpenAI is still working on another product that, for now, has no name. Allegedly, this product aims to help OpenAI customers launch new models safely.
Allegedly, some of OpenAI customers have already had access to this product but, at the moment, no more information is known about the Artificial Intelligence company's plans to launch this novelty.
Returning to GPT-6 Cyber, Fortune notes that this is the fourth security-focused Artificial Intelligence model that is launched by OpenAI in 2026 – something that makes it clear that this area has become increasingly important to those in this area.
Australia accuses OpenAI of having “infiltrated” on official website
The Australian Prime Minister, Anthony Albanese, today accused the giant of OpenAI artificial intelligence of having “infiltrated” the government health services website, expressing his “extreme concern”.
“ Today I spoke with OpenAI’s head Sam Altman to convey Australia’s extreme concern for this event.”Albanese declared to the press in the margins of the United Nations Annual General Assembly in New York.
Agent AI infiltrated public and internal archives of the statistical health service, specified the Prime Minister.
“ For now, nothing indicates that there has been access to personal information, but investigations continue”, said Albanese, classifying the situation as “inacceptable”.
State services are also investigating to determine whether other government entities may also have been affected, he added.