Evidence of US Powerlessness After Trump Expels Advanced Weapons Supplier
Jakarta, CNBC Indonesia – Anthropic cannot be immediately and completely blocked from Trump administration government agencies, despite having provoked the president’s considerable anger. The Pentagon has suddenly permitted the use of Anthropic’s AI tool until a phased discontinuation period spanning six months.
A Reuters report citing an internal Pentagon memo states that Anthropic’s AI tools may still be used if deemed important for national security. Use is permitted only in rare circumstances.
Additionally, Anthropic may continue to be used in activities that critically support national security operations and where no suitable alternatives exist, according to reports from Thursday, 12 March 2026.
Nevertheless, usage must still pass through an exemption approval process. Pentagon units seeking to obtain such exemptions must submit comprehensive risk mitigation plans.
The Pentagon has confirmed the existence of the internal memo but has declined to comment further. Anthropic has not yet responded to requests for comment.
This report indicates how difficult it is for the United States government to completely disengage from Anthropic. Franklin Turner, a government contracts attorney at McCarter & English, stated that the memo amounts to an acknowledgement of how difficult it is for certain vendors to ensure they have removed Anthropic from their entire supply chain.
“I expect there will be a great many exemption requests,” he said.
Previously, the relationship between Anthropic and the US government was quite close. Anthropic was a supplier of AI tools for US military defence needs and received a massive $200 billion contract in 2025.
However, the relationship between Anthropic and the US government became strained ahead of US-Israeli attacks on Iran. Anthropic refused to allow its AI tools to be used to develop autonomous weapons and to spy on US citizens.
The Trump administration became extremely angry and described Anthropic as a left-wing “woke” company that endangers American lives and national security. Trump subsequently banned the use of Anthropic at government agencies and added it to the supply chain risk blacklist.
Nevertheless, the Wall Street Journal reported that the US military continues to use Anthropic’s AI tools in operations against Iran. Anthropic’s Claude AI has been involved in intelligence research, target identification, and combat simulations.
Anthropic has refused to accept its placement on the supply chain risk blacklist and has taken the matter to court. Legal proceedings are ongoing.