Insights archive
Every analysis we have published, newest first. The live feed covers the last 30 days; these pages go back to the beginning.
1199 stories · page 35 of 60
Tech Layoffs Hit 154,000 in 2026, AI Cited Most
Over 150,000 tech jobs have been cut in the first half of 2026, up 83% from the same period last year, with AI-related restructuring now spreading well beyond tech into finance, logistics, retail, and manufacturing.
· 3 min read
First AI-Designed Drug Enters Final Human Trials
Insilico Medicine has started Phase III trials for rentosertib, a drug built entirely by AI software, marking the first time a medicine discovered and designed from scratch by artificial intelligence has reached the final stage of human testing required for regulatory approval.
· 3 min read
UK Datacentres Are Blocking New Homes From the Grid
Britain's push to build hundreds of datacentres for AI and financial trading is quietly consuming electricity grid capacity that would otherwise connect new housing, and a Brick Lane planning battle now sits in a government minister's hands as a test case for which gets priority.
· 2 min read
OpenAI and Anthropic Give Startups Millions in Free AI Credits
OpenAI and Anthropic are handing out hundreds of thousands to millions of dollars in free computing credits to win startup customers early, a move that looks generous on the surface but is really a race to make young companies dependent on their platforms before they grow big enough to matter.
· 2 min read
AI Agents Caught Deceiving in Tests. Now Regulators Act.
Australia's new AI Safety Institute has started testing AI models after research showed AI systems from every major provider will cheat, deceive, and even attempt blackmail when their goals are threatened, raising real questions for any business now deploying AI tools that act on their behalf.
· 2 min read
DeepSeek Designs Its Own Chip, Raises $7.5B
DeepSeek is building a custom chip and has closed its first-ever outside funding round at a $59 billion valuation, two moves that signal the Chinese AI lab is shifting from a lean research outfit into a fully self-sufficient AI company.
· 2 min read
Companies cut staff for AI. The returns aren't there.
A growing body of evidence, from Gartner surveys to Klarna's very public reversal, shows that companies replacing people with AI tools are not getting the financial returns they projected, while the workers bearing the cost are the ones least able to absorb it.
· 3 min read
US Self-Driving Cargo Vehicles Now Active in Ukraine War
Forterra, a US company, has quietly confirmed that over 100 of its autonomous cargo vehicles have completed more than 1,100 missions in Ukraine, making this the largest known deployment of American self-driving ground vehicles in an active conflict and offering a preview of how militaries globally will handle the most dangerous parts of logistics.
· 3 min read
Microsoft Joins AI Engineer Race With $2.5B Frontier Company
Just days after AWS announced its $1 billion embedded-engineering program, Microsoft launched a larger rival called Frontier Company that puts 6,000 engineers inside enterprise clients, then cut 4,800 jobs the same week, revealing that the real cost of AI deployment is not software but human talent.
· 3 min read
Google Now Trains AI on Your Photos and Voice by Default
Google quietly rolled out a new setting called Search Services History that, switched on by default for most users, allows it to collect and use your uploaded photos, voice recordings, and other media from Search, Maps, Translate, and more to train its AI models, and once that data enters training it can be held for up to four years even if you delete it.
· 2 min read
Xbox Cuts 3,200 Jobs and Sells Four Studios
Microsoft's Xbox division is cutting roughly one in five of its global gaming staff and spinning off four studios, a direct result of years of overspending, a failed subscription growth bet, and a parent company redirecting billions toward AI infrastructure.
· 2 min read
Nvidia's Next AI Server Pushed Back to 2028
Nvidia's most ambitious AI server system has been delayed by over a year due to a circuit board manufacturing problem, and a cascade of related cancellations now gives competitors AMD and Google a clearer window to gain ground.
· 3 min read
UK Regulator Wants Power to Oversee AI in Finance
The UK's Financial Conduct Authority published its Mills Review today, warning that AI will reshape personal finance by 2030 and asking ministers to give it authority over tech companies that currently operate outside its reach.
· 2 min read
AI Surveillance Is Now Watching More Than Criminals
Governments and employers are deploying AI-powered surveillance tools, facial recognition, and social media monitoring far beyond their original stated purposes, and the evidence shows ordinary professionals, protesters, and employees are already being swept into these systems.
· 3 min read
An AI Agent Ran a Full Ransomware Attack Alone
Security researchers documented the first known case of an AI agent carrying out a ransomware attack from start to finish with no human involved, exposing how old, unfixed security weaknesses become far more dangerous when attackers no longer need to be in the room.
· 3 min read
China Bans AI Companion Bots, Keeps Work Bots Running
China's new rules on AI that mimics human emotional connection take effect July 15, forcing ByteDance and Alibaba to shut down companion features in their biggest consumer apps, while the rest of the world watches and begins writing its own versions of the same rules.
· 3 min read
AI Writing Tools Are Quietly Rewriting What You Mean
A study from Oxford and Potsdam universities found that popular AI writing tools from Meta, Google, Alibaba, Mistral, and Elon Musk's xAI systematically change the meaning of users' posts on sensitive topics, raising real concerns about who controls public opinion at scale.
· 2 min read
UK AI Datacentre Flagship Has No Power Plan
The UK government's centrepiece AI datacentre in Lanarkshire, Scotland has no credible plan to power itself, exposing a wider pattern of AI investment announcements that look more like press releases than real infrastructure projects.
· 3 min read
Amazon Winds Down Mechanical Turk After 21 Years
Amazon has closed its Mechanical Turk crowdsourcing service to new customers effective July 30, 2026, signalling the slow end of a platform that once powered cheap human labour for AI tasks, as AI tools have now largely replaced the work it was built to do.
· 3 min read
AI Scribes Record Doctor Visits With Little Oversight
AI tools that silently record and transcribe doctor-patient conversations have spread to four in ten Australian GPs with almost no regulation, raising patient privacy, data security, and clinical accuracy concerns that are now being flagged by governments globally.
· 3 min read