Anthropic decision from Pentagon: “Risk to national security”

The US Department of Defense (Pentagon) made a remarkable assessment, stating that the use of artificial intelligence company Anthropic’s technologies in military infrastructure may pose “unacceptable risks”. In the official documents submitted to the court, reservations about the company’s integration with war technologies and defense systems were clearly expressed. Accordingly, Anthropic’s rejection of certain terms of use raised serious questions about the organization’s supply chain security. Although Anthropic argued that it acted in line with ethical concerns, the Pentagon stated that this approach could undermine operational reliability. These developments have reopened the boundaries of collaboration between artificial intelligence companies and government institutions.

According to the information in the court file, the US Department of Defense added a special provision to the contracts, demanding that artificial intelligence technologies can be used “for any legal purpose.” But Anthropic refused to allow its models to be used in mass surveillance or autonomous weapons systems. Despite this, the Pentagon stated that this attitude raises doubts about whether the company is a reliable business partner. On the other hand, while it was emphasized that artificial intelligence systems were open to manipulation, the possibility of Anthropic changing system behavior in line with its own “ethical boundaries” was evaluated as a risk factor. It was stated that this situation could have unpredictable consequences, especially during active military operations.

Anthropic opposed some use cases on ethical grounds

In the statements of the US Department of Defense, it was pointed out that artificial intelligence technologies play a critical role in the battlefield, and it was stated that these systems should be uninterrupted and controllable. In addition, the possibility of unilateral intervention by companies came to the fore as a factor that could cause serious disruptions in military planning. Despite this, Anthropic’s opposition to some usage scenarios on ethical grounds brought about a different debate in the technology industry. Because in recent years, many artificial intelligence companies have tended to impose restrictions on the use of their products for military and surveillance purposes. However, government agencies argue that such restrictions reduce operational flexibility.

In line with the Pentagon’s assessment, it was also included in the court documents that US President Donald Trump ordered federal agencies to stop the use of Anthropic technologies. The company opposed this decision and appealed to the judiciary and requested interim measures. In addition, it is stated that the “supply chain risk” label in question could lead to billions of dollars of revenue loss for the company. Although commercial activities other than defense projects can continue, it is stated that such a classification may also have an impact on the reputation in the sector.

On the other hand, it is known that companies such as Microsoft, Google and OpenAI, which are among the leading actors in the technology world, have offered opinions in support of Anthropic. This support reveals the common concerns of companies operating in the field of artificial intelligence. In addition, it is predicted that such disputes between the state and the private sector may arise more frequently in the future.