Taiwan says a hacking campaign linked to a foreign government hit its computer systems last month, and that the attackers used artificial intelligence tools to carry it out.
Meta CEO Mark Zuckerberg warned this week that the US could lose the AI race to China.
CFR's survey of foreign-policy experts puts the odds of coherent global AI governance at a median 29 out of 100. Meanwhile nation-state hackers are already folding generative models into every stage of the intrusion lifecycle, and Taiwan is racing to build an AI defense it cannot yet fund. Three threads, one convergence.
From Sunday, a European Union law called the AI Act requires companies to say clearly when a chatbot, image or video was made or altered by artificial intelligence.
The technology contest is moving from locks to rules — AI standards, data and industrial inputs — and Kenya, Mexico and the wider global South are where it is being decided.
For fifty years, powerful states kept dangerous technology behind a lock and rationed access as leverage. This week showed that reflex failing on three fronts at once — and what is replacing it.
A smart, AI-capable recording device was discovered in the Czech government's own meeting room at Strakova Academy — and stayed unreported in a drawer for weeks, with neither police nor intelligence services informed.