Insights archive
Every analysis we have published, newest first. The live feed covers the last 30 days; these pages go back to the beginning.
887 stories · page 23 of 45
Your ChatGPT History Can Be Subpoenaed in Court
A US federal arson trial used a suspect's ChatGPT conversation history as evidence, and while it ended in a mistrial, the legal pattern it confirms matters for every professional who uses AI chatbots at work or at home.
· 2 min read
Tesla Settles FSD Pedestrian Death Lawsuit
Tesla quietly settled a wrongful death lawsuit over its first known pedestrian fatality caused by its self-driving software, a case that is also driving a federal safety investigation that could force a recall of 3.2 million vehicles.
· 3 min read
Governments Worldwide Are Banning Kids From Social Media
Australia's December 2025 ban on social media for under-16s has triggered a chain reaction across more than 40 countries, while early evidence shows most children are still getting through, and courts have started holding platforms liable for addictive design.
· 2 min read
NYT Targets Microsoft Directly in OpenAI Copyright Case
The New York Times has moved to sharpen its lawsuit against Microsoft by arguing the tech giant did not merely fund OpenAI but actively built the computing infrastructure used to train AI models on stolen content, a distinction that matters because the Supreme Court just raised the legal bar for this kind of claim.
· 2 min read
Perplexity Launches AI Workflow Tool for Legal Teams
Perplexity has entered the legal software market with Computer for Counsel, a tool that automates the administrative side of legal work by connecting case law databases, document systems, and contract tools, and the timing matters because the legal AI market is crowding fast with big names raising billions.
· 2 min read
OpenAI Appoints First India Managing Director
OpenAI has hired Prabhjeet Singh, former head of Uber India, as its first managing director for India, the company's second-largest market with 100 million weekly ChatGPT users, as it races Google and Anthropic to convert massive free-user growth into actual business revenue.
· 2 min read
Stripe Cut Compliance Review Time 26% With AI Agents
Stripe built an AI-assisted system that cuts the time compliance staff spend on financial reviews by 26%, while keeping humans in charge of every final decision, offering a replicable model for any regulated business drowning in manual review work.
· 3 min read
California Launches First US AI Job Loss Tracker
California has built the first public dashboard in the US to track unemployment claims among workers in jobs most exposed to AI, offering a monthly data signal that any business operator can read, even though the data cannot definitively prove AI caused any specific layoff.
· 2 min read
Patronus AI Raises $50M to Test AI Agents Before They Fail
Patronus AI has raised $50 million to build simulated digital environments where AI agents are stress-tested before being deployed in real business workflows, a problem that is becoming urgent as agent failures inside companies are rising sharply.
· 2 min read
Gaming Clips Are Now AI Robot Training Data
General Intuition, a New York startup that trains AI agents using billions of video game clips, raised $320 million at a $2.3 billion valuation, with backing from Jeff Bezos and Eric Schmidt, on the premise that recorded human gameplay teaches machines to move through physical space better than any other data source.
· 3 min read
Amazon Raises India Cloud Investment to $48 Billion
Amazon has added another $13 billion to its India plans, bringing its total commitment to $48 billion by 2030, as the country pairs massive tech infrastructure spending with a rare tax holiday that lets foreign cloud companies operate nearly tax-free until 2047.
· 2 min read
UK Police Built a Crime-Prediction System That Mostly Got It Wrong
British police in Bristol secretly scored hundreds of thousands of people using AI risk models, some of which were abandoned after failing basic accuracy tests, just as the UK government launches a £75 million national program to spread AI tools across all 43 police forces.
· 3 min read
Netherlands Fights US Bill That Would Ban ASML China Sales
The US Congress is advancing a bill that would force ASML, Europe's most valuable company, to stop selling and servicing chip-making machines to China, prompting the Dutch government to send a minister to Washington to push back, with real financial and diplomatic stakes on both sides.
· 3 min read
Google Now Uses Your Photos and Voice Searches for AI Training
Google has started rolling out a new account setting called Search Services History that saves the images, audio, and video you use in Google Search and feeds them into AI model training by default, and most users will not know to turn it off.
· 2 min read
Ex-Infosys CEO Launches Startup to Replace IT Services Firms
Vishal Sikka, who once ran one of the world's biggest IT outsourcing firms, just raised $32 million to build a company that uses AI to do the work those firms charge hundreds of billions of dollars for, and he already has paying customers.
· 2 min read
Companies Are Rationing AI After Runaway Spending
After months of pushing staff to use AI as much as possible, large companies are now hitting unexpected bills with little to show for it, and the reckoning has arrived for any business operator still treating AI as a free-for-all.
· 2 min read
Google Adds Computer Control to Gemini 3.5 Flash
Google has built the ability for its Gemini 3.5 Flash AI to see and control a computer screen, browser, or mobile device, moving AI agents from answering questions toward doing actual office work on your behalf.
· 2 min read
Agility Robotics Goes Public via SPAC at $2.5B
Agility Robotics, which makes humanoid robots already working in real warehouses and factories, is going public through a merger deal that values it at $2.5 billion, making it the first standalone publicly listed humanoid robot company in the US with proven commercial deployments.
· 2 min read
OpenAI Launches Its First Custom AI Chip
OpenAI unveiled Jalapeño, its first purpose-built chip, made with Broadcom to cut the cost of answering billions of daily ChatGPT requests, as the company bleeds $20 billion a year in operating losses and prepares to go public.
· 3 min read
ChatGPT Gets Voice Mode That Holds a Real Conversation
OpenAI is preparing a new ChatGPT voice mode that can listen and speak at the same time, remember the full thread of a conversation, and handle interruptions naturally, which matters because it makes AI voice tools significantly more useful for any business that deals with customers or staff over the phone.
· 2 min read