The Anthropic Mythos 5 AI model has received limited approval from the United States government, allowing a select group of American cybersecurity companies to use the advanced artificial intelligence system after weeks of restrictions over national security concerns.
Anthropic announced on Friday that it had been authorised to provide access to Mythos 5 to a small number of US cybersecurity firms. The company said it remains in discussions with the Trump administration to further expand access to the model and eventually restore broader availability.
The decision marks a partial easing of restrictions imposed earlier this month after US authorities ordered Anthropic to suspend public access to two of its most advanced AI models, Mythos 5 and Fable 5.
Government Restricted Access Over Security Risks
On June 12, the US government directed Anthropic to immediately block access to its latest AI models after identifying vulnerabilities in the safety mechanisms designed to prevent misuse.
Officials expressed concerns that the models’ advanced capabilities could potentially be exploited in ways that threaten national security if sufficient safeguards were not in place.
The unprecedented intervention sparked debate across the technology industry, with critics arguing that the government’s action represented excessive interference in the development and deployment of artificial intelligence technologies.
Despite the criticism, US officials have maintained that advanced AI systems require stronger oversight as they become increasingly capable of performing sophisticated tasks that could have military, cybersecurity, or intelligence implications.
Progress Leads to Limited Approval
According to a letter from US Commerce Secretary Howard Lutnick, cited by Reuters and Politico, Anthropic has made significant progress in addressing government concerns.
“Anthropic has worked with the US government to address risks associated with the Covered Models. These efforts have yielded significant progress,” Lutnick said.
The Department of Commerce also defended its approach to regulating advanced AI systems.
Commerce Department spokesperson Benno Kass said officials have worked to ensure that the United States maintains its leadership in artificial intelligence while protecting national security interests.
The latest approval allows only a limited number of vetted cybersecurity organisations to use Mythos 5, rather than restoring full commercial access.
Anthropic said discussions with federal authorities will continue as the company seeks broader approval for Mythos 5 and the eventual return of its companion model, Fable 5.
Previous Dispute with the Trump Administration
The latest development comes after earlier tensions between Anthropic and the Trump administration over the company’s policies governing military use of artificial intelligence.
Anthropic previously declined to permit its technology to be used for mass surveillance or autonomous weapons programmes, a position that reportedly led the Pentagon to cancel contracts with the AI company.
The disagreement highlighted growing debates within the AI industry over how advanced models should be deployed in defence and national security applications.
Technology companies developing frontier AI systems continue to face increasing pressure to balance commercial innovation with ethical concerns surrounding military use, cybersecurity, misinformation and public safety.
OpenAI Also Faces New Restrictions
Anthropic’s announcement coincided with the launch of OpenAI’s latest flagship model, GPT-5.6, which is also being released under tighter government oversight.
OpenAI confirmed that access to GPT-5.6 will initially be restricted, with customers undergoing government-validated approval before receiving access.
OpenAI Chief Executive Sam Altman acknowledged that the approval process is more restrictive than the company would ideally prefer.
“This isn’t quite the process that we think is optimal,” Altman wrote on X.
However, he also expressed support for ongoing collaboration between the technology industry and government regulators.
“I believe the government shares most of our goals, and that they are overall doing a good job in a very difficult situation,” he added.
New Executive Order Expands AI Oversight
The restrictions on Anthropic and OpenAI come after President Donald Trump signed an executive order earlier this month establishing a voluntary federal review process for advanced AI models before they are released publicly.
The initiative is intended to identify potential national security risks associated with increasingly powerful artificial intelligence systems.
Although participation is officially voluntary, many leading AI developers are understood to be cooperating with federal authorities as they seek approval for new products.
The White House has provided limited details about how the review system will operate in practice or which future AI models will be subject to scrutiny.
Read Also: US Proposes New Tariffs on 60 Economies Over Forced Labour Concerns
Industry observers say the evolving regulatory framework reflects growing concerns among governments worldwide about ensuring that increasingly capable AI systems are developed responsibly without slowing technological innovation.
The latest approval for Anthropic’s Mythos 5 suggests that companies willing to strengthen safety protections and work closely with regulators may gain faster access to the market, even as broader debates over AI governance continue.
By DOTT TV Newsroom | AFP
