2024 in AI: It’s changed the world, but it’s not all good

The article surveys AI-related security problems in 2024, including breaches of AI platforms exposing user fantasies and 346,000 customer files, an information-stealer distributed via a fake free AI editor (Windows and macOS), and widespread use of deepfakes and voice cloning in high-value scams; it also notes warnings that AI could exacerbate ransomware threats while citing a benign countermeasure (AI Granny Daisy) that wastes scammers' time.