Anthropic’s effort to help companies boost their cyber-defences using its new AI model may irk not just hackers, but also America’s government https://www.economist.com/business/2026/04/08/how-dangerous-is-mythos-anthropics-new-ai-model?taid=69de2793bd318b0001e829b0&utm_campaign=trueanthem&utm_medium=social&utm_source=voante