Select Language:
OpenAI announced that its newest series of powerful AI models will be publicly launched this Thursday, following approval from the U.S. government for a wider deployment. This includes the latest GPT-5.6 models and other advanced AI tools, such as Anthropic’s Mythos series, which have raised concerns about their unmatched ability to detect vulnerabilities in code that hackers could exploit.
These capabilities have sparked national security worries. In late June, OpenAI disclosed that it had provided early access to GPT-5.6 to a select group of trusted U.S.-based partners, at Washington’s request.
Large language models form the backbone of chatbots and numerous AI applications, thanks to their ability to process enormous amounts of digital information. The GPT-5.6 lineup features three versions: Sol, the flagship model; Terra, a mid-range option suitable for routine tasks; and Luna, a fast, budget-friendly variant.
OpenAI stated on Tuesday via an X post that GPT-5.6 Sol, along with Terra and Luna, would be available to the public this week, with preview access expanding globally. While further specifics were not provided, Axios reported, citing a knowledgeable source, that the Trump administration had authorized a broad launch after testing phases and discussions with government officials.
OpenAI has been contacted by AFP for comment. This comes shortly after its competitor, Anthropic, the startup behind the Claude chatbot, also resumed offering access to its top-tier models, Fable 5 and Mythos 5, following the removal of restrictions by Washington.
During the previous administration under President Donald Trump, there was a push for fewer regulations on AI companies, aiming to give the U.S. an edge over China in artificial intelligence. Currently, the government is drafting new rules to determine which AI models will be subject to security restrictions, aligned with an executive order from the White House.
In June, OpenAI expressed concern about government access processes becoming a permanent standard, arguing that such measures could hinder users, businesses, and others who rely on these tools. The company also indicated it is collaborating with policymakers to develop a framework for future model releases under a cyber executive order.
Once widely accessible, Terra will be priced at roughly half the cost of its predecessor, GPT-5.5, as OpenAI seeks to secure market share amid stiff competition from Anthropic and Google. Both companies have filed confidential IPO paperwork with U.S. regulators and are eyeing public listings with valuations nearing $1 trillion, intensifying the competition in the AI industry.
ChatGPT
Perplexity
Gemini AI
Grok AI
