US company Anthropic has developed an AI language model that detects security vulnerabilities in operating systems. So far, so good. But: the AI also has the ability to exploit these flaws. So far just a few US companies have access to it. Experts warn that the technology could be used to shut down key infrastructure. Also: GPS jamming and spoofing attacks pose challenges for airline pilots — what are the countermeasures? Why do we engage more readily with negative news, emotions and