Anthropic has received significant relief from the US government. The company announced on Friday that select US cybersecurity firms have now been granted permission to use its advanced AI model, Mythos 5. This decision is viewed as a partial easing of restrictions that had been imposed due to national security concerns. The company clarified that the US administration had previously completely blocked access to Mythos 5, citing security reasons.
Anthropic stated that it would continue discussions with the government to extend access to Mythos 5 to other users as well. Additionally, the company aims to make its other AI model, Fable 5, available to the general public again. A company representative noted that talks are ongoing and the impact could become visible in the coming weeks.
This is not the first time a conflict has arisen between the US AI company Anthropic and the current US administration—specifically, the Trump administration. Relations between the two have been strained for several months. On June 12 of this month, the US government ordered Anthropic to immediately halt access to its two most advanced AI models after identifying flaws in their safety systems that could increase the risk of misuse. Many criticized this government move as an act of overreach—an instance of excessive interference.
As previously mentioned, a dispute had already occurred between Anthropic and the White House when the company refused to allow its technology to be used for mass surveillance and autonomous weapons. Following that decision, the Pentagon cancelled its contracts with Anthropic, causing relations between the company and the US government to deteriorate significantly. Restrictions Imposed on GPT-5.6 Model Too
It is worth noting that on Friday, June 26, Anthropic’s rival, OpenAI, also launched its new GPT-5.6 model; however, the government restricted access to it, and individual customer approvals are being verified at various government levels. This clearly indicates that the AI industry is currently under strict scrutiny by the US government.
Earlier this month, President Donald Trump signed an executive order requiring advanced AI models to undergo a voluntary federal review process prior to public release, in order to assess associated national security risks. The White House has not yet disclosed full details regarding how this order will be implemented, how companies will participate, or which specific AI models will be subject to this review framework.
