With the world’s major economies pouring investment into artificial intelligence, ensuring the technology’s safety before its applications are known has become an increasingly urgent policy imperative. While a sci-fi-style AI apocalypse is not impossible, more immediate risks to both security and democracy must be addressed.
