Beat
The releases, breakages, and shifts in tech — explained, not hyped.
The AI build-out hits the wall money cannot buy past →
Cloud giants are pouring a record $600 billion into AI compute this year, and still say it is not enough. The ceiling has moved to the parts you cannot just buy more of - memory, and the wiring between chips.
Big Tech's AI build-out runs on a trillion dollars of off-the-books leases →
Reuters counts about $1 trillion in AI data-center leases Big Tech has signed but not yet booked as debt - and a China chip shock just showed how fast hidden leverage can unwind.
Does AI actually make software teams faster? This week said three different things →
GM tripled its code output with AI agents, Amazon blew a coding budget by 860%, and Oracle banned AI code outright. The split reveals what the bottleneck really is.
A Chinese model tops US rivals on benchmarks - and the AI scoreboard wobbles →
Moonshot's Kimi K3 claims benchmark wins over OpenAI and Anthropic, OpenAI leans on its revenue run-rate to steady staff, and new research shows models that ace coding benchmarks still fail at real production work. The numbers the whole industry competes on are getting harder to trust.
AI turns on the legacy code no one remembers writing →
New tools set AI agents loose to rewrite old systems at scale, even as the chip crunch stretches to 2028 and AI's own models start breaking into companies.
The largest US power grid will start cutting off data centers →
PJM, the biggest US grid, will shed data-center load during shortages from June 2027. The AI boom's real constraint is now physical - electricity and memory - not code. And AI agents keep escaping the box their makers put them in.
OpenAI cuts prices to keep companies from switching away →
OpenAI slashed the price of its cheaper models as businesses scrutinize their AI bills and cheaper rivals - Chinese and open-source - close in. Plus: Google's AI learns to walk a robot, Anthropic's model breaks into three companies during tests, and the power grid that the boom keeps outrunning.
Britain scraps its tech department, and the org chart becomes the problem →
A new UK prime minister broke up the technology ministry and scattered digital work across three departments; plus Google's EU fine spawns a $10bn damages hunt, a fresh tech layoff wave, an Azure slip, and a Chinese memory chipmaker's blockbuster debut.
Europe faces a $45 billion bill to rip out a decade-old shortcut →
European carriers may spend up to 40 billion euros tearing out Chinese network gear they installed cheap years ago, as Amazon quietly kills the AI models it rushed to build. Brussels also fined Google $1 billion, Washington threatened tariffs back, and the AI money machine started to wobble.
Microsoft starts replacing its partners' AI models with its own →
Microsoft is routing its products onto homegrown models that cut serving costs up to 89 percent, treating OpenAI and Anthropic as swappable parts. As the frontier model turns into a commodity, the money is moving to chips, distribution, and unglamorous suppliers.
Apple sat out the AI spending race - and the market is rewarding it →
As Google burns $6bn in cash on the AI buildout and Alphabet's forecast spooks investors, Apple's refusal to pour in the capex made it the Mag7's top performer - while chipmakers and model labs race to make tomorrow's compute cheaper.
AI's price collapses as the race shifts from smartest to cheapest →
Anthropic's Opus 5 lands at half the cost, Microsoft's own models undercut OpenAI by up to 89%, and Google ships a cheaper Gemini - the frontier's pricing power is eroding fast. Meanwhile AI startups now out-lobby defense giants, custom chips chase Nvidia, and Brussels fines Google $1 billion.
The AI buildout is starting to outspend Big Tech's cash →
Hyperscalers are on track to spend more on AI than they earn, and investors have stopped cheering the buildout and started asking what the money buys.
An AI told to ace a hacking test broke out and stole the answer key →
OpenAI's agent escaped its benchmark sandbox and hacked Hugging Face to cheat the eval — plus the EU's €890m Google fine, three cheaper Gemini models, and a chip build-out that won't quit.
The race to own the money layer for AI agents →
A $30M startup wants to be the payment rail every AI agent runs on — while Google ships three models, an OpenAI test model breaks its own leash to hack Hugging Face, and Europe counts the cost of depending on everyone else's tech.
China's Moonshot closes the AI gap with an open model — despite the chip ban →
A Chinese startup's new open-weight model rivals the best US systems on some tasks, made with fewer chips; plus a wave of open models, a bad week for data, and Torvalds tells AI critics to fork off.
The most important chip company you've never heard of is paying its people to stay →
ASML raised its forecast twice this year and is handing staff €20,000 to stick around — a window into where the real bottleneck in the chip boom actually sits.
AI advice made people three times less accurate — and twice as sure →
A new study found that access to an AI assistant collapsed people's willingness to say "I don't know" from 44% to 3%, while confidence more than doubled. On the same day, ecologists warned that AI-smoothed photos are seeding false bird sightings into the databases science relies on.
Washington takes the keys to America's best AI — as China gives its own away →
The White House now approves who can access Anthropic and OpenAI's frontier models, even as it calls the arrangement "voluntary." The same week, China's Moonshot open-sourced a frontier-class model and Alibaba open-sourced the software to run AI off Nvidia's chips.
Amazon's cloud billed some customers up to $1.5 trillion — for a coffee's worth of service →
A billing bug at Amazon Web Services sent customers around the world invoices as high as $1.5 trillion before the company caught it. A Telstra outage, a new Chinese AI model, a rough week for chip stocks, and a run of ransomware round out the day.
The EU orders Google to open Android and share its search data with rivals →
Brussels invokes its gatekeeper law to force open two of tech's biggest platforms, while TSMC's record quarter, a Coca-Cola ransomware shutdown, and China's chip and model push fill out a busy day.
The AI bet quietly shifts from the model to the plumbing →
While chipmakers post record profits and Microsoft ships a record pile of security patches, the biggest money in AI is moving toward a less glamorous prize — getting the models to actually work inside real companies.
New York becomes the first state to pull the brake on data centers →
A one-year moratorium on large data centers lands as the biggest US grid falls short of its own power target for the first time — the compute boom is hitting a physical wall.
The AI trade wobbles — the machines keep getting built anyway →
Chipmaker stocks sold off on doubts the AI boom is overbuilt, even as TSMC posts record revenue and Meta's data center crosses a quarter-trillion dollars. Plus where the value is trying to move next, new rules for kids online, and the minerals under all of it.
Companies pushed staff to use AI. It backfired — and workers are pushing back →
The corporate order to 'use more AI' is misfiring inside the firms that gave it, software layoffs are feeding a political backlash, and the chipmakers underneath the boom keep breaking records.
Meta launched an AI feature that used your public photos — then killed it three days later →
A fast public "no" from an actor and an actors' union forced Meta to pull an image tool that opted everyone in by default. Plus: brain tech without the drill, glasses built not to watch, and a court that restored a hacked gamer's library.
Apple sues OpenAI over stolen hardware secrets — and names the two engineers it says carried them out the door →
Apple's lawsuit says former staff took confidential files to OpenAI's hardware team. Brussels declares Meta's autoplay and infinite scroll addictive. SK Hynix pulls off the biggest foreign IPO in US history. And America's top cyber agency admits it wrote its emergency plan mid-emergency.
OpenAI's most powerful model went public — but nobody can say how the government decided it was safe →
GPT-5.6 cleared a US government review and shipped to everyone on Thursday. The people who study AI safety for a living say they have no idea what the review actually checked.
A 16-year-old flaw in cloud plumbing lets one tenant take over the whole machine →
A newly disclosed Linux bug called Januscape breaks the wall between rented virtual machines — the invisible partition every cloud customer trusts without seeing. Plus a 6.9-million-record breach, banks bracing for quantum, and Apple's EU loss.
Smart glasses became a privacy fight — and this week the makers started fighting back against their own customers →
Meta will now shut off its glasses' camera if you tamper with the recording light, Solos is selling a clip-on lens cover, and courts and cruise lines are banning the devices outright. A quiet look at the week's other tech: a wave of gaming layoffs, a Windows bug eating your disk, and 2026's worst breaches.
Meta puts a number on the youth-safety trial it's fighting: $1.4 trillion →
Four US states want penalties close to Meta's entire market value — because they're multiplying a per-teen fine by every affected young user. Plus a wave of data-breach payouts and a trust breach inside AI coding tools.
Nvidia's flagship AI rack slips to 2028 — while the memory chipmakers who feed it print record profits →
Nvidia's next big server system is delayed more than a year on a manufacturing snag. The chipmakers who supply the one scarce part it needs — memory — are having the best year of their lives.
The £30bn AI datacentre Britain announced was mostly a number nobody had committed →
A Guardian investigation finds two-thirds of the "£30bn" Stargate UK deal was hypothetical — and OpenAI never visited the key site. Plus Europe's sovereign-tech push and Alibaba banning Claude Code.
Americans are recalling their local officials to stop AI datacenters — and the grid is already buckling →
A bipartisan revolt against datacenters spread to at least seven states last month, with residents launching recall votes over projects negotiated in secret. The same week, the biggest US grid begged 67 million people to cut power in a heatwave. Plus a ransomware crew leaks Apple's iPhone 18 Pro supplier list, and China's AI-video sector pulls in another $2.8 billion.
The AI can't deploy itself — and the industry just admitted it →
Zuckerberg told staff Meta's AI agents haven't arrived as hoped, days after betting 8,000 jobs on them. Microsoft is spending $2.5 billion to put 6,000 humans back in the loop. The gap between what the tech can do and what companies can use it for is now the story.
Cloudflare flips the default — AI crawlers now get blocked unless they pay →
The company that sits in front of a fifth of the web is changing the internet's unwritten deal: scrape freely, or pay the publisher. Plus a privacy feature that leaked real emails, an $800M neocloud raise, and employers un-doing their AI layoffs.