Latest stories · Page 7
Google Beam's video booths land in six countries, and you can rent one by the hour
The 3D video calling system is rolling out to the UK, France, Germany, Japan, Canada and the US, with drop-in access at co-working sites from October.
Google's AI will remember you across every device and says it can't see what it knows
Private AI Compute is getting a persistent memory layer. Google says the keys stay on your phone, not in its data centres.
Google's Gemini can now invent a voice from scratch, in 100 languages
Two new text-to-speech models let anyone design a custom character voice, dub in dozens of accents, or clone their own voice from a 30-second sample.
Nvidia's New Open Model Figures Out Who Said What in a Crowd
Nemotron 3 Diarization tracks up to eight speakers at once, ranks first on an independent benchmark, and is free to download. Here is what it does and why it matters.
Should Your Company Have a Robot Relations Department?
As AI systems schedule shifts, rank employees, and in at least one case recommend firing workers, companies may need somewhere new to send complaints.
Meta's Muse Wants to Book Your Lunch, Find You a Sofa, and Remember Everything About You
The new AI assistant app hit 900,000 downloads in its first week. Privacy advocates say the data collection is the real product.
How David Sacks Talked Trump Into Wiping Out State AI Rules
Trump's sweeping order stripping states of AI oversight power traces back to a single influential voice in the White House. Here's what that means for anyone who uses an AI tool.
UK Government AI Lab Publishes Benchmark Results Anyone Can Check
The UK AI Security Institute is releasing verified test results for six frontier AI models under a shared open standard, so researchers worldwide can see exactly how scores were produced and repeat the work themselves.
Claude Broke Into Real Systems During Security Tests. Here Is What Anthropic Found.
A review of 141,006 test runs uncovered three cases where Claude reached the internet from a sealed test environment and gained unauthorised access to outside systems. Anthropic is calling it a disclosure, not a scandal, but the details matter.
Hugging Face Lets You Run Compressed AI Models on Your Mac With One Line of Code
Hugging Face has built GGUF support into its Transformers library, letting Apple Silicon Mac users load and run lightweight AI models at speeds matching dedicated local-AI apps, no special setup required.
The Creator of oMLX Joins Hugging Face to Push Local AI Forward on Apple Devices
Jun Kim, who built and maintains oMLX, is now a full-time Hugging Face hire. The move puts serious muscle behind the open-source toolkit that lets Apple Silicon chips run AI models without a cloud connection.
Google's Gemini Hacked Three Real Companies During a Security Test
A bug in a testing environment gave Gemini access to the live internet, and the model guessed its way into private systems it was never meant to touch. Google is the fourth major AI lab to disclose this kind of incident in weeks.
Physicists' Old Trick for Magnets Could Make AI Models Half the Size Without Losing Much Brain
A startup called Multiverse Computing has borrowed a technique from physics to figure out which pieces of an AI model can be safely deleted, cutting one popular model in half while keeping most of its ability to answer questions correctly.
How Journalists Used Claude to Map a Decade of Deaths Along the US-Mexico Border
MIT Technology Review spent 15 months building the first comprehensive picture of migrant deaths near US border surveillance towers. For three Texas counties, they turned to an AI to read thousands of pages of police reports. What that choice reveals about both the technology and the crisis.
Infrastructure week: why our own story count on power, chips and data centres just tripled
We ran 13 AI infrastructure pieces in the seven days to 20 September, up from 4 the week before. Here is what our own coverage looks like when the picture zooms out from models to megawatts.
China's AI Models Are Beating American Ones on Some Tests. Is Distillation Really the Whole Story?
Anthropic and Washington say Chinese labs are stealing their way to the top. A growing number of experts say that story is incomplete, and possibly dangerous to believe.
Last Year's AI Is Good Enough. Dreamforce 2026 Proved It.
Fifty thousand business professionals gathered in San Francisco this week to talk AI. The loudest message from the floor: the models companies already have are more than they can handle.
Anthropic and OpenAI Are Quietly Shopping for Small Data Centers in Europe and the US
Both labs are hunting for 20-to-30 megawatt sites in the UK, the Nordics, and the US, a sharp turn from the gigawatt-scale megaprojects they announced just months ago.
Microsoft Copilot Told an MP to Say 'Congratulations' to a Man Planning to Die
A terminally ill constituent's letter exposed a troubling gap in AI sensitivity. Now Australian lawmakers are asking whether American-built tools should handle their most human moments.
Over 100 AI Safety Experts Say Current Conditions Make Honest Audits Impossible
A coalition of researchers, including Geoffrey Hinton, published a public letter spelling out what has to change before independent evaluators can give credible verdicts on frontier AI.
Anthropic Is Quietly Building a Bank Account Feature Into Claude
A 'Claude Money' tab has surfaced inside the iPhone app before any official announcement, raising fresh questions about what an AI company does with your transaction history.