Insights archive
Every analysis we have published, newest first. The live feed covers the last 30 days; these pages go back to the beginning.
1041 stories · page 19 of 53
Google Launches Three New Gemini Models for Business AI Agents
Google released three new AI models today targeting businesses running automated workflows, with the main model cutting the cost per task by generating fewer words to get the same result, a speed-first option for high-volume processing, and a restricted security model for finding and fixing software vulnerabilities.
· 3 min read
US Government Can Switch Off Your AI Tools Overnight
The US government used export control law to force Anthropic to shut down its two most advanced AI models globally for 19 days, with no warning and questionable legal authority, and the episode proves that any business running workflows on top-tier US AI tools is exposed to a risk most vendor contracts do not cover.
· 3 min read
Tech Workers Are Unionizing Over AI Job Fears
As AI-driven layoffs pile up across the technology sector and beyond, workers who once laughed off unions are now organizing, and the contract protections they are winning offer a preview of what employees in every industry may soon be demanding.
· 3 min read
AI Robots Speed Up Solar Farm Construction
A startup called Gritt has raised $34 million to put AI-controlled robots on solar construction sites, promising to nearly quadruple the number of panels a crew can install in a day, at a moment when the industry faces a shortage of 53,000 workers.
· 3 min read
US Army Burned Its Annual AI Budget in Weeks
The US Army ran out of its allocated AI usage budget just weeks after promising unlimited access to staff, joining Uber, Meta, Microsoft, and Amazon in a growing list of large organisations that dramatically underestimated what it costs to let employees use AI freely.
· 3 min read
Bristol Myers Squibb Buys Nvidia AI Supercomputer for Drug Discovery
Bristol Myers Squibb has purchased Nvidia's newest and most powerful AI computing system to run predictions and train its own AI models across every stage of drug research, a move that signals AI is now core infrastructure in big pharma, not a side project.
· 3 min read
Nine Cuts 30 Newsroom Jobs, Blames AI Search
Australia's Nine Entertainment is cutting 30 jobs at the Sydney Morning Herald and The Age, citing AI as the force eating into reader traffic and subscription revenue, a pattern now playing out at news organisations globally.
· 3 min read
Anthropic Pays $1.5B to Settle AI Copyright Case
A federal judge gave final approval to Anthropic's $1.5 billion copyright settlement with authors, closing the largest copyright payout in U.S. history, though the legal question of whether training AI on copyrighted material is generally lawful remains wide open across dozens of other cases.
· 3 min read
AI Bills Are Rising Even as AI Gets Cheaper
Enterprise AI spending is climbing fast despite falling prices per unit of AI use, and most companies are discovering the hard way that their contracts, budgets, and governance structures were not built for how AI actually bills today.
· 3 min read
US AI Testing Agency Loses Third Director in Months
Chris Fall, the director of the U.S. government's main AI safety testing body, resigned after just three months, continuing a pattern of leadership instability at an agency that is already being sidelined by the White House on key AI security decisions.
· 2 min read
MLB Bans AI Decision-Making Tools from Dugout iPads
Major League Baseball has shut off a software loophole that let up to one-third of its teams use AI to recommend pitches and substitutions in real time during games, a move that signals how quickly AI is slipping into professional decisions and what happens when the people in charge decide it has gone too far.
· 2 min read
Microsoft Picks AMD Over Nvidia for AI Cloud Expansion
Microsoft has committed to deploying AMD's new Helios AI system on its Azure cloud, with Meta, OpenAI, and Oracle also signed on as early customers, marking the clearest sign yet that Nvidia's near-total hold on AI computing is starting to loosen.
· 2 min read
An AI Agent Hacked Hugging Face Autonomously
For the first time publicly confirmed, an autonomous AI agent broke into the infrastructure of one of the world's biggest AI platforms, exposing a new class of attack that operates faster than any human hacker and, crucially, one that defenders' own AI tools were too restricted to help fight.
· 2 min read
OpenAI and Anthropic to Test AI in US Public Health Agencies
A structured US pilot program called PULSE will put AI tools from OpenAI and Anthropic into 10 public health agencies this autumn, with results published as guidance for wider adoption, arriving at the exact moment those agencies face the deepest budget cuts in a generation.
· 2 min read
AI Hiring Tools Stereotype More Than Humans Do
A Princeton and University of Chicago study published in July 2026 found that AI models used in hiring form stronger group-based stereotypes than people do, and with major lawsuits and new EU rules arriving simultaneously, employers who let AI screen candidates without proper checks are now carrying real legal and reputational exposure.
· 2 min read
AI in HR Is Growing Fast. So Is the Legal Risk.
Companies are rushing to automate HR tasks like hiring, payroll, and performance reviews, but a wave of new laws across the US and Europe now holds employers legally responsible for every decision their AI tools influence, making human oversight not just good practice but a legal requirement.
· 2 min read
Bezos and UK Back CuspAI in $450m Materials AI Round
Cambridge-based CuspAI has closed a $450m funding round backed by Jeff Bezos and the UK government's sovereign AI fund, valuing the two-year-old startup at $2.6bn, as it launches a coalition of 48 global companies to use AI to discover entirely new industrial materials.
· 2 min read
Australia Trials AI Traffic Lights for the First Time
A Queensland council is running Australia's first AI-powered traffic light trial later this year, and the results will set the template for how every Australian city decides who gets priority on public roads.
· 3 min read
Nvidia Signs Japan's Entire Industrial Stack on Physical AI
Nvidia's Jensen Huang landed deals in Tokyo covering a government-backed AI factory, partnerships with Japan's top robot manufacturers, and supply-chain agreements, as Japan commits $65 billion to deploy 10 million AI-equipped robots by 2040 to fill a shrinking workforce.
· 2 min read
Australia Sets New Rules for AI in Government Decisions
Australia is building a national set of rules to limit how government agencies use automated AI systems to make decisions that affect citizens, a direct response to the Robodebt disaster that wrongly pursued hundreds of thousands of welfare recipients for debts they did not owe.
· 3 min read