Uncertainty Surrounds Use of Claude AI Model in Iranian School Strike

A deadly missile strike that killed around 120 children at an Iranian elementary school has raised questions about the use of increasingly capable artificial intelligence systems during military operations. The incident, which occurred in late February, is shrouded in uncertainty regarding the role played by the Claude AI model developed by Anthropic.

Anthropic's CEO, Dario Amodei, expressed his company's lack of knowledge about how their AI was used in the strike when speaking to Bloomberg on Wednesday. 'We don't have access to, we don't know exactly how these models were used,' he said, highlighting the complexities surrounding the use of advanced AI systems.

The incident has sparked concerns that military leaders may be relying too heavily on AI without fully understanding its capabilities or limitations. Amodei acknowledged this risk but emphasized that his company's policies are designed to prevent such misuse. 'The principle we have established is a human makes the final decision,' he noted, suggesting that Anthropic's guidelines were followed in this instance.

The Pentagon has not claimed responsibility for the strike, despite preliminary findings from U.S. military investigators indicating American forces may be responsible. An unnamed U.S. official told The Wall Street Journal that U.S. military investigators believe it is likely that American forces carried out the bombing.

Anthropic's AI models have been extensively used by the Defense Department and other national security agencies for various purposes, including intelligence analysis and operational planning. However, the company has faced controversy in recent months after refusing a request from Defense Secretary Pete Hegseth to grant unrestricted access to their AI technology. Amodei cited concerns about mass domestic surveillance and fully autonomous weapons on the battlefield.

The U.S. military subsequently scrapped its contract with Anthropic and designated the company as a supply chain risk to national security, which Anthropic is currently contesting in court.