AI News
AI news that matters. Updated daily.
No stories match your filters.
Anthropic Briefed Trump Administration on Mythos Before Any Public Announcement
The Trump administration got an early look at Anthropic's newest model before the general public did. An Anthropic co-founder confirmed the company briefed White House officials on Mythos, a model the company says has substantial cybersecurity capabilities, according to TechCrunch.
Two Attacks on Sam Altman in 48 Hours Signal a New Kind of AI Backlash
Two attacks on Sam Altman's home in 48 hours. A 20-year-old charged federally for throwing a Molotov cocktail at the OpenAI CEO's house, then attempting to break into OpenAI's headquarters. Then, two days later, a second apparent targeting of the same address.
Google Chrome's 'Skills' Feature Saves and Replays Your Favorite AI Prompts
Most browser AI tools treat every interaction as a one-off. You type a prompt, get a result, and repeat the same prompt tomorrow. Google is changing that with Skills, a new Chrome feature that lets you save favorite Gemini prompts and reuse them across any website.
Chrome's New 'Skills' Feature Turns Saved Prompts Into One-Click Browser Tools
Anyone who uses AI tools daily knows the drill: open the tool, paste your go-to prompt, tweak it slightly, run it. Repeat 20 times a day. Google just shipped a Chrome feature called Skills that turns that repetitive loop into a single click.
Chrome Adds Reusable AI Prompt "Skills" to Gemini on Desktop
Google Chrome now lets you save your favorite Gemini prompts as reusable "Skills" and fire them across any open tabs with a click. The feature is rolling out on Chrome desktop, and it addresses something every regular AI user runs into: the same handful of prompts get retyped from memory dozens of times a week.
How to Set Up Chrome's AI Skills: Premade Options and Building Your Own
Chrome's new Skills feature in the Gemini sidebar comes with several premade options ready to use: a YouTube video summarizer, a recipe protein optimizer, and additional presets for common browsing tasks.
Claude Code Adds 'Routines' for Automating Repeated Development Instructions
What do you type at the start of every Claude Code session? For many developers, it's the same block of preferences: preferred test framework, naming conventions, how errors should be handled. Claude Code's new Routines feature offers a way to save and reuse those instruction sets by name.
Anthropic and OpenAI Split on Illinois Bill That Would Limit AI Liability
The two biggest AI companies publicly agree on a lot - federal AI regulation, safety commitments, the need for oversight. They have now found something they disagree on: how much legal protection AI labs should have when their products cause mass harm.
Google Gemini Personal Intelligence Rolls Out in India with Gmail and Photos Access
Google has extended Gemini's Personal Intelligence feature to India, one of Android's largest markets. The feature connects your Google accounts - Gmail, Photos, and other services - so Gemini can draw on personal context when answering questions rather than working from general knowledge alone.
Google Disputes Claim That Its SynthID AI Watermark Was Reverse-Engineered
Google's SynthID system is designed to do one job: invisibly mark AI-generated images so they can be identified later, even after basic editing. A developer going by the username Aloshdenny says they've figured out how the system works from the outside - and published code on GitHub that, they claim, can both strip those watermarks from AI-generated images and inject them into regular photographs to make human-made images look AI-generated. Google says that's not what happened.
Stanford AI Index: Adoption Hits Record Pace as China Closes Gap with U.S.
What does it look like when a technology spreads faster than any previous wave in computing history? Stanford's annual AI Index report says we're in it. The report describes AI adoption's current pace as "historic" - and flags China's progress against the U.S. as a finding that deserves more attention than it's getting from most Western coverage.
The Hidden Cost of Letting AI Write Code You Never Read
Last year, vibe coding felt like a cheat code. Describe what you want, accept the AI's output, ship the feature. This year, the bills are coming due.
OpenAI Acquires Personal Finance Startup Hiro to Add Financial Planning to ChatGPT
OpenAI acquired Hiro, an AI personal finance startup, in a move that signals what's coming next inside ChatGPT: help with your money.
Federal Charges Filed Against Man Who Attacked Sam Altman's Home and OpenAI HQ
Daniel Moreno-Gama, 20, is facing federal charges after allegedly driving from Texas to California with the stated intent to kill OpenAI CEO Sam Altman. On April 10th, he was arrested after throwing a Molotov cocktail at Altman's San Francisco home and then attempting to force his way into OpenAI's headquarters.
Anthropic's Safety Trust Now Holds Board Majority After Novartis CEO Appointment
For the first time, Anthropic's independent oversight body controls the majority of the company's board. The Long-Term Benefit Trust (LTBT) - a group of directors with no financial stake in Anthropic, designed to keep the company accountable to its stated mission - has appointed Novartis CEO Vas Narasimhan as a new board member, pushing Trust-appointed directors past the 50% threshold.