A new study found that code generated by AI is more likely to contain made-up information that can be used to trick software into interacting with malicious code.

A new crowd-trained way to develop LLMs over the internet could shake up the AI industry with a giant 100 billion-parameter model later this year.

WhatsApp’s AI tools will use a new “Private Processing” system designed to allow cloud access without letting Meta or anyone else see end-to-end encrypted chats. But experts still see risks.

Liking features on social media can provide troves of data about human behavior to AI models. But as AI gets smarter, will it be able to know users’ preferences before they do?

Italian essayist Andrea Colamedici tells WIRED Hypnocracy: Trump, Musk, and the New Architecture of Reality was a “philosophical experiment and a performance.” The book’s Chinese author does not exist.

Poorly maintained sewers can have disastrous consequences, but regular inspections can be time-consuming, expensive, and dangerous. The solution: subterranean dung drones.

Looking for the CliffsNotes of a lengthy YouTube video? This Gemini feature could be worth a try.

Social Security workers are being asked to use an AI chatbot. An animated video on how to do so failed to mention that the chatbot can’t be trusted with personally identifiable information.

Perplexity? Gemini? Copilot? The Razr Ultra will have it all optimized for its flippy, folding screen.

Google’s AI Overviews feature credible-sounding explanations for completely made-up idioms.