OpenAI is partnering with the defence-tech manufacturer Anduril Industries to develop military technology for the Pentagon, according to a joint statement from both. The ChatGPT owner’s AI models will be used to detect and shoot down drones that threaten US and allied forces, which the companies say will “reduce the burden on human operators, and improve situational awareness”. The deal marks a step change for OpenAI, which earlier this year barred anyone from using its models for “weapons development” or “military and warfare”. It also signifies a changing attitude in Silicon Valley, where military contracts were once taboo. In 2018, Google turned down a $10 billion defence contract after 4,000 employees signed a petition in protest. Now companies including Anthropic, Meta and Google are openly stepping up collaboration with the Pentagon.