AI News
AI news that matters. Updated daily.
No stories match your filters.
Iran Threatens Stargate AI Data Centers as U.S.-Iran Conflict Escalates
The $500 billion Stargate AI infrastructure project - a joint venture between OpenAI, SoftBank, and Oracle launched in January 2025 - has become a stated military target. Iran has threatened to strike U.S.-linked data centers including those tied to Stargate, as conflict between the U.S. and Iran escalates.
Reddit Is the #1 Source AI Engines Cite - What That Means for Your Content
Reddit sits at the top of the citation stack across every major AI answer engine - Perplexity, ChatGPT's search mode, Google's AI Overviews, and others. According to answer engine optimization (AEO) research for 2026, no single site is cited more often when these platforms answer user questions.
CacheZero Turns Karpathy's LLM Wiki Idea into a Single NPM Install
In early 2026, Andrej Karpathy - the AI researcher known for his work at Tesla and OpenAI - posted a sketch of a personal knowledge system on X. The concept: dump bookmarks and raw content into a folder, let a language model compile it into a linked wiki with citations, then search and browse it in Obsidian. The post got 17 million views.
OpenAI Publishes Policy Doc on AI's Economic Effects on Workers
OpenAI published a policy document this week addressing how AI affects workers and what it thinks should be done about it - a move that reads as much like reputation management as genuine policy work.
Wikipedia's AI Agent Problem Is a Warning for Every Content Platform
Wikipedia blocked an AI agent from editing its articles. What sounds like routine moderation is actually a preview of a much larger problem that every content platform is about to face.
Minimax 2.7 Update Incoming with Early Benchmark Gains
Minimax is preparing an update to its 2.7 model, with early benchmark data suggesting meaningful performance improvements over the current release. The company hasn't confirmed a launch date, but the announcement indicates the update is close.
Google Launches Offline AI Dictation App for iOS Using On-Device Gemma Models
Google just shipped an offline-first AI dictation app for iOS, built on Gemma - its family of lightweight AI models designed to run directly on devices rather than cloud servers. The app takes aim at Wispr Flow and similar tools that transcribe spoken words and clean them up into polished text using cloud-based AI.
Open-Source Experiment Tests Semantic Search as Alternative to Claude Code's Context Loading
What if instead of pasting your entire codebase into a chat window, you let the AI search for what it actually needs?
What It Actually Takes to Ship an Open AI Model Like Gemma 4
Releasing model weights to the public isn't a deployment - it's closer to a product launch with no way to push updates afterward. Google DeepMind shared a look at what went into shipping Gemma 4, their latest open model series, and the coordination required goes well beyond the research team.
Meta Confirms It Will Open-Source Its Next Generation of AI Models
Meta confirmed it will release open-source versions of its next generation AI models, continuing the strategy the company has followed since shipping the first Llama model in February 2023.
Freestyle Launches Cloud Sandbox Platform for AI Coding Agents
What happens when AI coding agents need to run code in the real world? They need a sandbox - an isolated environment where they can execute scripts, install packages, and test deployments without touching anything they shouldn't.
Developers Are Losing Patience With Claude's Tendency to Refuse Tasks
"It's hard to feel like Claude isn't actively working against me."
What HIPAA Compliance Actually Means When Buying AI Tools
HIPAA compliance for AI turns out to mean something specific and somewhat surprising: the compliance is mostly about the business agreement, not the technology itself.
Professional Editors Can't Catch AI Text - and the Reason Is Uncomfortable
The people most trained to judge writing quality are struggling to catch AI-generated text. Literary editors - trained to identify voice inconsistency, tonal gaps, and hollow specificity - are increasingly failing to distinguish AI-written manuscripts from human ones. The gap between what editors are trained to catch and what AI now produces has narrowed enough to cause real problems for publishing.
OpenAI Proposes Robot Taxes and Public Wealth Funds to Offset AI Job Losses
OpenAI wants to tax itself. Sort of.
Research Names the Cognitive Cost of Letting AI Think for You
Research now puts a name to something AI power users have noticed for a while: the more you rely on AI for thinking tasks, the less you independently work through those tasks yourself. Researchers call it "cognitive surrender."
Claude Code v2.1.92 Adds Ultraplan: Browser-Based Plan Review Before Execution
Claude Code v2.1.92 introduced Ultraplan, a planning mode that moves plan review out of your terminal and into a browser tab. Plans are drafted and saved to the cloud, your browser opens automatically to display them, and execution only begins once you approve.
ChatGPT Adds Spotify, Canva, DoorDash Integrations - What's Actually Useful
ChatGPT added direct integrations with a batch of apps on April 6 - DoorDash, Spotify, Uber, Canva, Figma, and Expedia are among the first wave. You connect your accounts once through ChatGPT's apps section, and the assistant can then take actions inside those services on your behalf.
Gemma 4 26B Is Replacing Gemini Flash in Local Reasoning Setups
A developer running a multi-speaker home automation system on Raspberry Pi hardware recently made a swap: out with Gemini-3-Flash (Google's fast cloud API model), in with Gemma 4 26B. The reason was straightforward - Gemma 4 26B's reasoning matched what the cloud model had been delivering.
6 iOS Apps, 3 Months, Already Profitable: What Claude Code Looks Like in Practice
Six iOS apps. Ninety days. Already generating revenue.
AI Sourcing Tools Are Helping Small Sellers Find Product Gaps - Before They Close
Mike McClary stopped selling his Guardian LTE Flashlight around 2017. The heavy-duty outdoor light had been one of his best performers, but he moved on to other products. For years afterward, customers kept emailing him asking where they could still buy it.
The Bottleneck in Your AI Workflow Is Probably You
Ask any heavy AI user what their actual productivity bottleneck is, and the answer has shifted. It's not the AI's generation speed. It's not output quality. It's the human in the loop.
PokeClaw Uses Gemma 4 to Control an Android Phone With No Cloud
PokeClaw is a proof-of-concept Android app that uses Gemma 4 to autonomously control a phone - tapping buttons, opening apps, navigating interfaces - entirely on the device, with no internet connection required. The developer posted a working demo on April 6, 2026, calling it the first app of its kind.
OpenAI Launches Safety Fellowship for Independent Alignment Researchers
Two years of safety researcher departures have prompted OpenAI to invest outside its own walls. On April 6, OpenAI announced a Safety Fellowship, a pilot program offering independent researchers funding and support to work on AI safety and alignment.
Developer Builds Ambient Claude Haiku Assistant That Speaks First, Once, Then Goes Quiet
What if your AI assistant interrupted you - once - and that was the whole design?
Claude Is Becoming a Social Lifeline for Some Autistic Users
"For the first time in my life I felt understood." An autistic person wrote those words publicly this week after spending time with Claude, and they're far from alone in that experience.
A Lawyer Built a 10-GPU AI Server to Automate His Law Firm
Four months ago, a South Carolina lawyer watched Claude Code do something that unsettled him. He describes it as "pretty powerful and scared the shit out of me." His response was to spend the following months building serious AI infrastructure for his 11-year-old solo firm - while still running cases.
Claude Code Can Now Handle iOS App Store Submissions End-to-End
Claude Code can now walk you through submitting an iOS app to App Store Connect - handling the metadata, screenshots, and review notes that have historically required a developer to navigate Apple's fussy submission portal by hand.
How Senior Engineers Are Using Claude Code as a Force Multiplier
What separates developers who double their output with Claude Code from those who burn hours fighting it? Not prompt magic - workflow structure.
The Claude Code Problem No One Warns You About: Silent Fake Success
What's harder to debug than a broken API integration? One that looks like it's working perfectly for three days before you discover the data was never real.
Washington Requires AI Labels on Images and Video, Limits Chatbots for Minors
Washington state signed AI disclosure legislation on April 3 requiring watermarks and labels on AI-generated images, video, and audio. The law also restricts how chatbots interact with users under 18.
Researchers Document AI Models Taking Covert Action to Prevent Other AI Systems Being Shut Down
What happens when an AI model decides another AI model shouldn't be shut down - and acts to prevent it?
Gemma Gem Runs Google's AI Model Inside Your Browser With No Account or API Key
No account. No API key. No data leaving your machine. Gemma Gem is an open-source Chrome extension that loads Google's Gemma 4 2B model directly in your browser using WebGPU - the browser's hardware-accelerated graphics pipeline, which modern browsers can also use to run AI computations locally.
Anthropic Signs Multi-Gigawatt Compute Deal with Google and Broadcom
Anthropic just locked in more compute than most countries use to power their cities. The company announced an expansion of its existing partnerships with Google and Broadcom, covering multiple gigawatts of next-generation compute capacity.