A federal court in San Francisco has ruled that the Pentagon's blacklisting of American artificial intelligence company Anthropic earlier this year was unlawful, CNBC reported, BTA reported.
The US Department of Defense violated the First Amendment to the US Constitution by designating Anthropic as a supply chain risk "based on a desire to publicly rebuke" the company, according to the ruling by US District Judge Rita Lin.
Lin added that while the government is owed deference in matters related to national security, its actions were not based on any "clearly articulated basis“.
“The defendants argue that because of “Antropic“s “increasingly hostile behavior in the media“ and its criticism of the Department of Defense's views on the use of artificial intelligence, the defendants “cannot trust “Antropic“ to ensure the integrity of its models“, Lin wrote in his ruling.
“Neither the Constitution nor the federal law relied on by the defendants authorizes them to impose sweeping sanctions based primarily on “Antropic“s criticism of the administration's views“, the ruling states.
In March, the Department of Defense designated “Antropic“ a supply chain risk, suggesting that the company threatened U.S. national security, after as negotiations over how the military could use its AI models, Claude, spiraled out of control. The company demanded assurances that its technology would not be used to create autonomous weapons or mass surveillance of citizens, and the Department of Defense wanted Antropic to give the department unrestricted access to the Cloud for any legitimate purpose. The negotiations escalated, then collapsed, and Antropic became the first U.S. company to be publicly designated a supply chain risk, preventing defense contractors from using the developer's technology in their work with the department. Antropic has filed lawsuits against President Donald Trump's administration in San Francisco and Washington, D.C., in an attempt to revoked its blacklisting.
The company's lawsuit in Washington, D.C., is still ongoing.