Anthropic CEO Dario Amodei is negotiating with the US defense department to prevent the company from being designated a supply chain risk.The dispute arose after the firm refused to allow its Claude AI to be used for mass surveillance or the operation of lethal autonomous weapons.Despite the friction, the US military utilized the software to coordinate air strikes on over a thousand targets in Iran during a recent operation.Concurrently, hackers exploited AI chatbots to steal 195 million identities from Mexican government agencies in a massive data breach.