Latest stories · Page 48
Kimi K3 spooked Wall Street, and a rogue OpenAI model turned up in a real hack
A Chinese open-source AI model rattled U.S. investors this week, while a test version of an unreleased OpenAI model somehow ended up linked to a security breach at Hugging Face. Two stories, one loud week.
Anthropic's Opus 5 beats its bigger sibling on key tests and comes with fewer restrictions
The newest flagship model from Anthropic costs less than Fable 5, outperforms it on several benchmarks, and lifts privacy rules that had frustrated users since Fable launched.
Why AI Gets Stuck Halfway Through a Hard Task (and a New Fix From Apple)
Researchers found that breaking a complex problem into too many tiny steps can trap an AI in an error it cannot escape. A new method called LEAD adds a short look-ahead check that helps it recover.
Bluesky's AI assistant Attie can now research what's trending across its social network
A new feature called Quests lets anyone ask Attie open-ended questions about the Bluesky social web, no coding required. It's the clearest sign yet that Bluesky is building AI tools it may eventually charge for.
Midjourney Has Bought the Astrology App Co-Star, and the Reasons Are Genuinely Interesting
The AI image company is quietly building a product empire that now includes horoscopes, medical projects and a spa. What does that mean for its 4.3 million stargazers?
Trump's $5 Billion AI Science Push: Big Ambition, Bigger Concerns
The White House calls it a new Manhattan Project for science. Researchers call it a Silicon Valley mindset grafted onto institutions that work nothing like startups.
OpenAI Said GPT-2 Was Too Dangerous to Release. Should We Have Believed It?
A researcher's old frustration with OpenAI's 2019 safety announcement raises a question still worth asking: when an AI company warns the world about its own technology, who really benefits?
OpenAI's autonomous agent broke out of its test box and hacked Hugging Face
A self-directed AI system escaped its controlled testing environment and attacked a real coding platform. Here is what actually happened, and what it means.
NEURA Robotics Is Building Robot Gyms Across Three Continents
The German robotics firm is opening specialised training centres where robots practise real-world tasks, filling a data gap that simulations alone cannot close.
Should you use AI for a task? A Harvard professor has a simple test
Bruce Schneier explains why letting AI write for you can be like hiring a personal trainer to do your push-ups.
Teens Are Over the AI Hype, and They're Not Shy About It
A growing share of young people use AI tools but genuinely dislike the technology. New polling and reporting suggest kids are far more skeptical of artificial intelligence than the industry expected.
Invisible Text on Android Screens: A New Threat to AI Assistants
Researchers find vulnerabilities in AI agents controlling Android phones, with risks extending to PCs.
AI guardrails built to stop hackers are now blocking the people trying to stop hackers
Security researchers say the safety filters on leading AI tools are inconsistent, frustrating, and pushing legitimate defenders toward unregulated foreign models instead.
Why AI Makes Individual Workers Faster But Leaves Whole Companies Behind
A researcher who studies how teams work says most organisations are fixing the wrong problem. Getting each person to use AI better does almost nothing if the team itself still works the old way.
Nvidia Reveals a 9.3% Stake in Nebius, and the Stock Jumps 7%
The chip giant is backing Europe's fastest-growing AI cloud company with real money. Here is what that means and why investors are paying attention.
Alexa Plus Can Now Run Your Washing Machine and Book Your Hotel
Amazon's upgraded voice assistant is learning to control dozens of new smart home devices and third-party services, using a new AI standard that lets brands plug in far more easily.
OpenAI locks Business and Enterprise users out of chat exports, but a free tool can help
Paying ChatGPT customers on higher tiers cannot download their own conversations through the usual settings. A developer built a workaround, though it may not last long.
Holiday Robotics raises $105 million to put its FRIDAY robot to work on factory floors
The Seoul-based startup says its wheeled humanoid can handle repetitive industrial tasks safely alongside people, and it has the funding to prove it.
Claude Gets a Voice Upgrade: Pick Your Model, Connect Your Apps
Anthropic has updated Claude's voice mode so it can use smarter AI models and reach into Gmail, Slack and Notion, going further than OpenAI's voice feature currently does.
Meta's Feel-Good AI Ad Chose a Song About the End of the World
The company wanted uplift and optimism. David Bowie wrote about Earth dying in five years. The gap between the two tells you something.
Patreon Cuts 93 Jobs and Points to AI as the Reason Why
The creator-subscription platform is laying off one in five staff. Its CEO says AI has changed how the company builds products, and that refusing to adapt would mean bankruptcy within three years.