OpenAI said that it plans to give a select group of government-approved partners early access to GPT-5.6 Sol, its most powerful AI model to date, before releasing the product more broadly. The limited rollout follows a request from the Trump administration, which asked the company to adopt a phased launch strategy for its next-generation AI system while security reviews are conducted. 

Last week’s request came from the White House’s Office of the National Cyber Director and Office of Science and Technology Policy, which have pushed AI developers to give federal agencies early access to frontier models so officials can evaluate their capabilities and potential security risks before wider deployment,

OpenAI CEO Sam Altman told employees in a memo that GPT 5.6-Sol will initially be available to 20 approved partners, including Amazon’s Bedrock platform. According to the memo, access is being granted on a “customer-by-customer” basis while the review process is underway. 

“We’ve made clear to the U.S. government that this is not our preferred long term model, and will work with them and others in industry to achieve a more sustainable approach for future releases,” Altman said in the memo. Altman said he hopes to release GPT-5.6 to the public a “couple of weeks later.”

The decision reflects growing concern over what the Trump administration says are legitimate national security implications of increasingly capable AI systems. 

OpenAI says GPT-5.6 Sol is its most advanced model to date, with improvements in reasoning, autonomous task execution, software engineering and cybersecurity-related capabilities. It released benchmarks that said its performance was broadly comparable to Anthropic’s Mythos 5, which was withdrawn on June 12 following a directive from the Commerce Department expressing concerns that its advanced capabilities could create new cybersecurity threats. 

At the time, Anthropic said the concerns were over “a small number of previously known, minor vulnerabilities” and that “other publicly-available models are able to discover them as well without requiring a bypass.” 

Politico reported that the initial vulnerability was brought directly to the White House by Amazon CEO Andy Jassy. (Amazon is an investor in Anthropic.)

“We have reviewed a report that we believe is the basis of the government’s directive and validated that the level of capability displayed there is widely available from other models (including OpenAI’s GPT-5.5), and is used every day by the defenders who keep systems safe,” Anthropic said in its statement.

Anthropic and the Trump administration have been sparring for months since a dispute over how Anthropic’s models would be used by the Pentagon. Reports that the company’s models were used in U.S. operations in Venezuela in January, in violation of its licensing terms, which prohibit using Claude models to commit violence, led to a near complete rupture. 

Adding further layers to the drama, just hours after OpenAI announced the limited rollout of GPT-5.6, Anthropic disclosed that the Trump administration had approved a limited release of Mythos 5, reversing the Commerce Department restriction. 

The restrictions on GPT-5.6 Sol and Mythos 5 follow a recent call from the Five Eyes alliance for closer coordination on advanced AI development and security. A White House official said the administration continues “to collaborate with frontier AI labs to develop shared approaches for addressing the challenges of scaling this technology.” 

The post Trump administration allows limited GPT-5.6 release appeared first on The Media Copilot.

NewsGuard, a company best known for rating the reliability of online news sources, on Tuesday launched NewsGuard AI, a chatbot that draws exclusively from a database of journalist-vetted stories instead of the open web.

The launch comes as concerns persist over the accuracy of AI-generated responses. NewsGuard said a yearlong audit of leading AI models found they repeated false or misleading claims on controversial news topics 35% of the time. The company argues that limited responses to vetted sources can help reduce the spread of misinformation through AI systems. 

NewsGuard AI attributes information directly to the publishers whose reporting is used in its responses, unlike other chatbots like ChatGPT, Claude, Gemini, or Perplexity.

Participanting publishers include The Atlantic and other regional newspapers, opinion journals, and public media organizations. Readers, subscribers and members of some participating outlets will receive a free trial of NewsGuard AI followed by an offer for 33% off the chatbot’s standard $6 monthly subscription. 

The company also says it will share revenue with participating publishers through a 50-50 revenue-sharing model and affiliate-style subscription referrals, though it has not publicly disclosed the formula used to calculate payouts.

NewsGuard says its journalists have reviewed more than 36,000 sources since 2018, including newspapers, magazines, opinion publications, local news outlets, independent newsletters, government websites, think tanks, hospitals and research universities. Of these, roughly 12,000 have been rated reliable and are eligible to be cited by NewsGuard AI. 

The new service enters a rapidly evolving market in which publishers are negotiating licensing agreements with AI companies while also challenging the unauthorized use of their reporting. Media organizations have struck content deals with companies including OpenAI, Amazon and Meta, even as lawsuits and public disputes over AI scraping continue across the industry.

Chris Richmond, CEO of the fact-checking website Snopes, said the arrangement addressed concerns his organization has had with other AI products.

“Snopes has restricted most AI chatbots from scraping our content,” Richmond said. “But we’re happy to partner with NewsGuard on a model that does this the right way.” 

In addition to drawing from vetted sources, NewsGuard AI says it incorporates 41 editorial safeguards. These include access to NewsGuard’s database of 64,000 debunked false claims circulating online, which the company says help prevent the chatbot from repeating known misinformation. Users can also access detailed explanations debunking false claims and share them with others. 

“Few things will matter more in the near future than the ability of humans to figure out what’s real, what’s false, and what’s confabulated nonsense,” said Nicholas Thompson, CEO of The Atlantic. “This is particularly true when it comes to news.”

NewsGuard is also targeting educational institutions. Students at participating schools and universities will receive free access while enrolled. The company says the chatbot has been designed to refuse requests to write essays or reports for users. 

“NewsGuard AI can provide reliable research while not substituting for students doing their own writing and thinking,” said NewsGuard’s Chief Operating Officer Matt Skibinski.

Local language versions of NewsGuard AI will be available in French, German and Italian in September. 

The post Can AI deliver trustworthy news? NewsGuard thinks its new Chatbot has the answer appeared first on The Media Copilot.