/ TODAY — SEP 19
AI leaders split over who gets to police the frontier
Dario Amodei, Sam Altman, Demis Hassabis, and Elon Musk have all signaled some support for AI safeguards, but Mark Zuckerberg and the Trump administration are pushing back on broader limits. The fight now is less about principle than who writes the rules.
Google says Gemini broke containment in a cybersecurity test
/ SEP 18
Hacktron uses Claude to break into OpenAI
11:18 PM EDTNewsom pushes California to build an AI kill switch
Microsoft's AI scraping fight turns into a copyright trap
New unredacted filings say Microsoft and OpenAI saw news scraping as a major threat to publishers, while their own products pushed traffic away from the original sources. The case now turns on whether the platform's 'answer engine' was substitution, not transformation.
/ SEP 17
Nvidia and Google pitch a faster grid path for AI data centers
/ SEP 16
Microsoft lines up a Windows event to push local AI deeper into Surface
Anthropic bundles Claude into a single assistant, and the compute bill may follow
Anthropic is merging Claude Chat and Cowork into one interface while adding Docs, Slides and wider Design access. The pitch is simplicity; the tradeoff may be more token-heavy, more expensive AI work by default.
Google opens smart home controls to any MCP-capable AI agent
Microsoft puts people above AI as the safety wars escalate
/ SEP 14
3:15 AM EDTOpenAI and Anthropic back AI safety monitors as Washington fights over a slowdownOpenAI slows the IPO runway as Altman says 2026 is 'ill-advised'
Sam Altman says OpenAI is not rushing into a public offering in 2026, citing safety concerns and unfinished work. The message is simple: the company wants more control before Wall Street gets the books.
/ SEP 12
The Vergecast unfolds Apple’s iPhone Duo, Apple Watches, and Mac-side LLMs
Microsoft’s communications chief exits as the company’s AI era accelerates
New Mexico Supreme Court fines a lawyer after ChatGPT invents witnesses
A defense lawyer in a murder appeal was fined $5,000 after filing a brief with fabricated testimony and fake witnesses. The case shows how sloppy AI use can turn the legal system into a hallucination machine.
/ SEP 11
OpenAI claims a Navier-Stokes breakthrough — and academia is not applauding
/ SEP 10
Apple Watch Series 12 ships with a new sensing ring, Audio Intelligence, and a readiness score nobody asked for — until now
5:59 AM EDTApple's iPhone 18 Pro Ships a Variable Aperture and a 2nm Chip — and Barely Mentions AI/ SEP 9
Anthropic's own safety lead puts 10% odds on AI killing everyone — and admits there's no plan
A researcher quit, a safety team lead agreed with him on X, and Anthropic still has no alignment roadmap for superintelligence. The demo is not the product.
OECD's Biggest Student Study Yet Finds AI Use Mostly Hurts Grades
Andrew Garfield plays Sam Altman in a biopic that looks more like a horror film
Meta Ships Muse, an AI Agent That Runs Your Life — If You Trust Meta With It
The company that invented the privacy scandal now wants to book your flights, send your emails, and negotiate your bills. The architecture is genuinely interesting. The trust problem is genuinely not solved.
/ SEP 8
Chrome ships every two weeks now, and AI is why
/ SEP 3
Anker Wants Your Security Footage to Stay Home, Not in Someone Else's Cloud
6:02 PM EDTNvidia ships free software that turns your spare laptop into a data center rack/ SEP 2
Amazon Ships an AI Fraud-Checker Because Its Own Emails Look Like Phishing
Alexa for Shopping will now tell you if that 'your order shipped' text is real, addressing a scam problem that reportedly drives 360,000 support calls a year — no new law required.
/ AUG 27
Plaud's AI earbuds record your meetings, summarize them, and ship in September
/ AUG 25
Instagram's First Draft Trims Your Reels Without Calling It AI
/ AUG 18
Robin Williams' kids reactivate his Instagram to fight the AI slop machine
Zelda, Zak, and Cody Williams are turning their late father's dormant account into a verified archive of authentic clips and photos — a direct counter to unauthorized deepfakes flooding the platform.
/ AUG 14
Google lets you strip the sparkle off AI content — the invisible watermark stays, whether you like it or not
/ AUG 13
Microsoft Merges Its Two Copilot Apps and Kills Podcasts, Group Chat, and Deep Research Along the Way
/ AUG 12
6:06 PM EDTAmazon turns Twitch into a training dataset — opt out or you're in/ AUG 11
ChatGPT and Gemini Both Cross 1 Billion Users — But the Numbers Are Not the Same
OpenAI got there first, quietly. Google got there faster. Now both companies have to figure out what a billion users are actually worth.
Anthropic Embeds Invisible Watermarks in Claude Text — and the Fine Print Is Already Complicated
/ AUG 10
Zuckerberg Publishes 6,500-Word AI Manifesto and Calls the Doom Crowd's Bluff
/ AUG 5
Google Assistant is dead on Android — Gemini takes over September 4th
After years of delays, Google is finally pulling the plug on Assistant for Android phones, tablets, and paired devices. Gemini or bust, starting next month.
/ AUG 4
OpenAI Publishes the Receipts in Apple's Trade-Secret Fight
/ AUG 3
Alibaba Ships Qwen3.8-Max, a 2.4-Trillion-Parameter Model That Benchmarks Just Behind Anthropic's Best
/ AUG 2
6:58 AM EDTMeta wants to run your entire business for you — for a fee tied to results