Anthropic Sued the Trump Administration After Being Dubbed a National Security Risk

JAKARTA - Artificial intelligence company Anthropic sued the United States government after being labeled a controversial "supply chain risk" by the Pentagon. This lawsuit marks an escalation of an open conflict between technology companies and the government of US President Donald Trump regarding the use of artificial intelligence in military operations and domestic surveillance in the country.

In a lawsuit filed in federal court in California on Monday, March 9, Anthropic stated that the government's decision violated the law and was contrary to the principles of freedom of expression and due process rights guaranteed by the United States Constitution.

"This action is unprecedented and unconstitutional. The Constitution does not allow the government to use its enormous power to punish a company for protected views," Anthropic said in a lawsuit.

The "supply chain risk" label given by the Pentagon last week effectively restricts the use of Anthropic's AI technology in defense projects. The decision was made after the company refused to remove restrictions on the use of its AI chatbot, Claude, especially for two sensitive things: fully autonomous weapons and mass surveillance of American citizens.

US Defense Secretary Pete Hegseth is said to have set the risk assessment. The government also announced a six-month transition period before the full cessation of the use of Anthropic technology in Pentagon projects.

Anthropic also asked the court to overturn President Trump's order banning federal employees from using the Claude chatbot in their work.

This dispute sparked a wide debate about the limits of using artificial intelligence in modern warfare. On the one hand, the US government demands full flexibility to use AI in all uses that are considered legal by law. On the other hand, Anthropic believes that the technology is not yet reliable enough to be fully handed over to autonomous weapons systems.

According to the company, the use of AI to control weapons without human intervention has the potential to pose a major risk.

"Even the best AI models today are not yet sufficiently reliable for use in fully autonomous weapon systems," said Anthropic.

The Pentagon declined to comment on the ongoing legal process. But defense officials have previously insisted that decisions about how to defend the country should not be determined by private companies.

"U.S. laws, not technology companies, determine how this country is defended," the Pentagon official said.

What makes this case even hotter is the fact that the "supply chain risk" label is usually used to protect national security systems from foreign companies or entities that are considered potentially harmful. This is the first time the US government has used the mechanism against domestic technology companies.

Amid the conflict with the government, Anthropic is still open to negotiations. The company insists that the legal action does not mean closing the door to dialogue with the US government.

"We do not want to conflict with the government of the United States and remain open to reaching a settlement," the company official said.

This feud also dragged in AI industry rivals, especially OpenAI. Hours after the Pentagon sanctioned Anthropic, OpenAI reportedly signed a cooperation agreement with the Pentagon for the use of its technology in defense projects.

The case is expected to set an important precedent for the global artificial intelligence industry, particularly regarding how technology companies negotiate ethical limits on the use of AI in the military and state surveillance.

In terms of business, the risk label can have a significant impact on Anthropic. Although the company's CEO, Dario Amodei, assessed the impact to be relatively limited, most of the company's revenue comes from the business sector and non-military government agencies.

Anthropic is projected to earn around $14 billion this year, with more than 500 customers each paying at least $1 million a year to use Claude's technology in computer programming and other digital tasks.

Behind this legal conflict lies a fairly large philosophical question: who should determine the moral limits of super-advanced technologies such as AI - the government holding state power, or the company that created the technology?