Insights archive
Every analysis we have published, newest first. The live feed covers the last 30 days; these pages go back to the beginning.
1036 stories · page 25 of 52
New Zealand Towns Are Wary of AI Datacentre Deals
New Zealand's government is courting up to NZ$30 billion in foreign investment to build AI datacentres, but the first approved project in the small town of Makarewa is already drawing community pushback over water use, power consumption, and a jobs promise that global evidence suggests will not hold up long-term.
· 3 min read
OpenAI Folds Atlas Browser Into ChatGPT Desktop App
OpenAI is shutting down its nine-month-old Atlas browser and moving its AI web capabilities into a Chrome extension and a redesigned ChatGPT desktop app, a signal that for now the major AI companies have largely concluded that the browser itself is not the prize.
· 2 min read
AI Meeting Notetakers Carry Real Legal and Data Risks
AI notetaking bots are now common in video calls, but the way they collect, store, and sometimes share your voice and words creates legal exposure and data risks that most businesses have not thought through.
· 2 min read
Nvidia Backs Paris Voice AI Startup Gradium at $100M
Gradium, a Paris-based voice AI startup, has raised $100 million in total seed funding with Nvidia joining as a new investor, as the race to build AI that can hold a natural phone conversation without awkward pauses turns into one of the fastest-growing and best-funded corners of the AI industry.
· 2 min read
SAP Cuts Maintenance Fees After EU Pressure
The European Commission forced SAP to loosen its grip on support fees for on-premises business software, giving companies more freedom to use cheaper third-party providers, just as SAP pushes them toward its cloud AI products.
· 3 min read
Google Now Labels AI-Made Ads on Search and YouTube
Google has added an AI disclosure label to ads across Search, YouTube, and Discover, but the system only works automatically when Google's own tools were used to make the ad, leaving a large gap for third-party AI-made content.
· 2 min read
Meta Plans to Sell Its Spare AI Computing Power
Meta is building a cloud business to sell excess AI computing capacity to outside companies, a move that signals the AI infrastructure market is shifting from a pure race to acquire resources toward a race to generate returns from them.
· 3 min read
NYT Accuses OpenAI of Hiding Evidence in Copyright Case
News organizations led by The New York Times are asking a federal judge to punish OpenAI for allegedly lying about its ability to search its own data for proof that ChatGPT reproduced their journalism, a legal move that could strip OpenAI of its main defense and expose the broader AI industry to significant licensing costs.
· 2 min read
Estonia Uses AI to Check Laws After Tax Typo Cost $28M
A single wrong word in Estonia's gambling tax law left online casinos untaxed for months, and the resulting embarrassment pushed the government to build an AI tool that scans draft laws for errors, now rating 102 of 112 bills as high-risk.
· 3 min read
Anthropic Ends Flat-Rate Access to Its Best AI Model
Starting July 12, Anthropic's Claude Fable 5, its most capable publicly available AI model, moves to pay-per-use billing on top of existing subscriptions, marking the first time a major AI company has charged consumers extra beyond their monthly fee to access a flagship model.
· 2 min read
Colorado Softens AI Law, Federal Fight Continues
Colorado replaced its ambitious AI accountability law with a lighter disclosure-only version in May 2026, but a federal lawsuit backed by the DOJ and a White House campaign against state AI laws mean the new version could face legal challenges too.
· 3 min read
Meta Starts Making Its Own AI Chips in September
Meta begins production of its own purpose-built AI chips this September, part of a strategy to reduce its dependence on Nvidia as the company prepares to spend up to $145 billion on AI infrastructure in 2026, with direct consequences for how well-targeted ads across Facebook, Instagram, and WhatsApp will become.
· 2 min read
SambaNova Raises $1 Billion at $11 Billion Valuation
SambaNova, a chip startup that was nearly sold to Intel for $1.6 billion just seven months ago, has raised $1 billion at a $11 billion valuation, with JPMorgan Chase as its first major bank customer, signaling that large enterprises in regulated industries are beginning to install private AI hardware inside their own buildings rather than using the cloud.
· 2 min read
Ollama Raises $65M to Run AI Locally, Free of Cloud Fees
Ollama, a tool that lets developers run AI models on their own computers instead of paying per-use cloud fees, has raised $65 million and now sits inside 85% of Fortune 500 companies, signalling that businesses of all sizes are actively looking for cheaper, more private alternatives to the big AI services.
· 2 min read
OpenAI Sends Engineers to Work Inside Your Business
OpenAI's deployment arm has acquired Northslope, its second purchase in two months, as part of a $4 billion push to send hundreds of engineers directly into client companies to build and run AI systems, putting it in direct competition with consulting firms like Accenture.
· 2 min read
OpenAI Launches ChatGPT Work to Run Tasks for You
OpenAI launched ChatGPT Work today, a tool that connects to your existing apps, runs multi-step work tasks for hours without you watching, and delivers finished documents, slides, and reports.
· 3 min read
US Regulators Order Self-Driving Car Firms to Fix Emergency Response Problem
America's road safety regulator has told self-driving car companies they have until the end of July to present solutions after documenting a clear pattern of driverless vehicles blocking ambulances, freezing at fire scenes, and failing to recognize basic emergency signals.
· 2 min read
NHS Adds AI Blood Test to Womb Cancer Checks
Two NHS hospital trusts in England are rolling out a £30 AI-powered blood test that can identify women at very low risk of womb cancer before they go through uncomfortable internal scans, potentially sparing around 18,000 women a year from an invasive procedure.
· 3 min read
Meta AI Glasses Add a Privacy Fix, Then Undercut It
Meta updated its AI glasses to disable the camera when the recording light is physically removed, a real fix for a documented abuse problem, but on the same day it rolled out features that pull more of your personal data into its systems, and facial recognition code is already sitting on tens of millions of devices waiting to be switched on.
· 3 min read
Prime Intellect Raises $130M to Build Your Own AI
Prime Intellect, a startup that lets companies train and own their own AI agents without relying on OpenAI or Anthropic, has raised $130 million at a $1 billion valuation, arriving at exactly the moment enterprises are rethinking what it means to hand their operations over to a third-party AI provider they cannot control.
· 3 min read