Google Chrome is installing a 4 GB AI model onto your device. Here’s how you can turn it off
Google Chrome is silently downloading a 4 GB Gemini Nano model to your device without consent. Here's what it is, where it lives on your computer, and what you can do about it.
ChatGPT for Excel and Google Sheets is now out of beta, and it’s kind of a big deal
ChatGPT for Excel and Google Sheets is out of beta and now available globally. Build, edit, and ask questions about your spreadsheets using plain language.
Microsoft Copilot Cowork goes mobile and it’s ready to take tasks off your plate
Microsoft Copilot Cowork is now on iOS and Android, and it's adding reusable skills and third-party plugin support to help you delegate more work from anywhere.
Your ChatGPT history is a personality test you didn’t know you were taking
ETH Zurich researchers collected 62,090 real ChatGPT conversations from 668 users and trained an AI to predict personality traits from them. The results are a privacy wake-up call.
Xbox CEO scraps Copilot AI for consoles to refocus the platform on gameplay-first experiences
Microsoft has canceled Xbox Copilot AI development for consoles and mobile, as CEO Asha Sharma shifts focus toward player experience and core platform improvements.
iOS 27 could let users pick an AI model of their choice for text and image tasks
Apple's iOS 27 will reportedly let users swap AI models, including Google's Gemini and Anthropic's Claude, across Writing Tools, Siri, and Image Playground.
You can now trust Perplexity with medical tips, or at least a tad more than ChatGPT or Gemini
Perplexity launches Premium Health Sources, bringing trusted medical journals and clinical data into AI responses to improve accuracy and credibility.
Meta’s AI scans photos for bone structure to catch underage users on Instagram and Facebook
Meta is using AI visual analysis to scan photos and videos for physical indicators like height and bone structure to detect and remove users under 13 from Instagram and Facebook.
OpenAI could launch its first AI agent smartphone in 2027
OpenAI may be developing an AI-first smartphone targeting 2027, focusing on on-device intelligence, advanced processors, and next-gen memory to enable smarter, more autonomous user interactions.
Online ads are snitching more about your private life than you bargained for, finds research
New research shows that the ads you see online carry enough information for AI to reconstruct a detailed personal profile, without you clicking anything or sharing any data.
OpenAI is going Hollywood with ‘Critterz,’ and its Cannes-bound film used AI across its production
Critterz is being positioned as the first mainstream AI-assisted animated family film, arriving at Cannes amid a growing Hollywood debate over artificial intelligence.
AI chatbots continue feeding into our worst delusions, finds worrying report on ChatGPT and Grok
A new report details disturbing cases involving ChatGPT and Grok, adding to the growing concerns that AI chatbots can validate delusions instead of pulling users back.
Grok is about to join ChatGPT and Perplexity on your CarPlay dashboard
A placeholder in the latest Grok iPhone app confirms xAI is bringing Grok Voice mode to Apple CarPlay, making it the third AI chatbot to join the platform.
Google finally explains why Android AICore keeps eating your storage — and it actually makes a lot of sense
If you've ever noticed Android AICore taking up more storage than expected, Google now has an answer — and it turns out the culprit is a perfectly reasonable fail-safe you'd probably want anyway.
Yes, you should probably be nicer to your AI — here’s why that’s not as ridiculous as it sounds
You've probably felt it — that slightly hollow, going-through-the-motions quality when an AI interaction goes flat. Science is starting to explain why that happens, and who's responsible for it.
Academy just said it out loud: AI can’t win an Oscar for acting and writing
The Academy confirms AI cannot win Oscars for acting or writing, reinforcing that human performance and authorship remain central despite growing use of AI in filmmaking.
I put Gemini in charge of my Gmail, and it was eye-opening
I didn’t stop worrying about privacy, I just stopped letting it slow me down. And honestly, my inbox has never felt lighter.
AI got bougie? Research finds access skewed towards the rich, risking a new social divide
A new study finds AI awareness and usage are higher among wealthier groups, raising concerns about a growing digital divide and unequal access to opportunities.
Space data centers sound like a pipe dream. What if we put them on lamp posts?
A UK firm has signed a deal with Nigeria's Katsina State to deploy 50,000 solar-powered smart lamp posts that collectively function as a distributed AI data center, without using electricity from the grid.
ChatGPT just landed ads, Now, Google won’t rule out ads in Gemini app, of course.
Ads have started appearing inside ChatGPT, and Google is already leaving room for Gemini to follow. The chatbot ad era may be closer than users expected.
You can now protect your ChatGPT account with a special USB-key
OpenAI's new Advanced Account Security lets you ditch passwords entirely and lock your ChatGPT account with a physical USB key — and it's available right now.
ChatGPT is going to stop talking about goblins and gremlins as it sheds”nerdy” persona
OpenAI traces ChatGPT's goblin obsession to a "Nerdy" personality training error that spread across its models.
Your car’s Google Assistant is going away, and this is what replaces it
Google is bringing Gemini to cars, promising more natural conversations and smarter in-car assistance that actually understands what you mean.
Gemini can now turn your chat into a finished PDF, Word document, or spreadsheet in one tap
Google updated the Gemini app on April 29, 2026 to generate finished, downloadable files on request — nine formats including PDF, Word, Excel, LaTeX, and Markdown.
Your GM car just got a lot smarter with Google Gemini upgrade
Google Gemini is rolling out to 4 million GM vehicles, turning your daily drive into a smarter, more connected, and surprisingly fun experience.
Perplexity Comet Browser finally learns how to multitask on iPad
Perplexity Comet now works more naturally with iPadOS, giving users a stronger reason to try the AI browser over Safari or Chrome.
Apple is finally building the AI Photo editor that Google and Samsung have had for years
Bloomberg's Mark Gurman reports iOS 27, iPadOS 27, and macOS 27 will add three AI photo editing tools, Extend, Enhance, and Reframe, to the Photos app.
ChatGPT Images 2.0 is a game changer, but after testing it, the risks terrify me
For years, AI image generation has been framed as a story of steady progress. Every few months, the images look sharper, the prompts get easier, and the tools become more accessible. But that narrative is beginning to shift. The conversation is no longer just about quality or speed. It is about capability, intent, and where […]
Gemini wants to read your emails, calendar, and notifications to help you before you even ask
A Google app beta teardown reveals Gemini's upcoming Proactive Assistance feature, delivering personalized suggestions from user-selected apps like Gmail and Calendar.
Adobe Firefly AI is now live publicly, hoping you’ll talk to an AI and get work done
The Firefly AI Assistant public beta is available for Creative Cloud Pro and paid Firefly plan subscribers, with complimentary daily generative credits provided to eligible users during the beta period.
The next iPhone moment might come from an AI company, not Samsung or Apple
OpenAI is developing its own smartphone chip with MediaTek and Qualcomm, with Luxshare handling manufacturing. Mass production starts in 2028, and it could redefine how we use phones.
Discord users breach access controls to reach Anthropic’s Mythos model
Unauthorized access to Anthropic’s Mythos AI highlights growing concerns around safeguarding powerful systems, exposing vulnerabilities not in the model itself but in its surrounding access ecosystem.
I never thought AI would add typos – but it kind of makes sense
A new AI tool deliberately adds typos to emails, reflecting a shift where imperfect writing is seen as more human and trustworthy in the age of AI-generated communication.
NotebookLM can now automatically organize your research sources for you
Google has rolled out automatic source labeling and categorization to NotebookLM. The feature activates automatically once you have five or more sources in a notebook.
The best trick AI can pull is disappear into my gadgets instead of turning into a product
AI may finally become more useful when it stops acting like a product and starts quietly improving the gadgets people already own.
DeepSeek V4 is out, touting some disruptive wins over Gemini, ChatGPT, and Claude
DeepSeek V4 launched April 24, 2026, with two open-source models that rival closed-source leaders on coding and agentic benchmarks, at up to 9x lower API costs than OpenAI and Anthropic.
Sony’s table tennis robot made me think about what happens when AI gets a body
Sony’s table tennis robot looks like a lab flex with a paddle. The real story starts when AI stops answering prompts and learns to move through our world.
Scientists pretended to be delusional in AI chats. Grok and Gemini encouraged them.
esearchers tested five major AI chatbots with a simulated user showing signs of psychosis. Some made things worse. Others told the user to log off and call someone.