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 27 of 45
HPE and NVIDIA Launch Private AI Infrastructure for Agents
HPE and NVIDIA are expanding their joint on-premises AI system to support AI agents that can act autonomously inside companies, adding new hardware, security controls, and governance tools at a moment when most enterprises are deploying agents faster than they can manage them.
· 2 min read
Mobileye Launches Its Own Robotaxi Service in US
Mobileye, the Israeli company whose self-driving technology sits inside over 230 million cars globally, is stepping out from behind the scenes to run its own driverless taxi service in the US starting in 2027, entering a market already dominated by Waymo.
· 2 min read
ChatGPT Falls Below 50% Market Share for First Time
ChatGPT has dropped from dominating more than 80% of AI assistant users two years ago to below 50% today, as Google Gemini and Anthropic's Claude gain ground, and the race to turn AI users into paying customers is now as important as the race for raw numbers.
· 2 min read
Large Insurers Now Using AI Directly in Underwriting Decisions
A new independent benchmark of 30 major insurers finds that AI has moved from back-office automation into the core of risk decisions, with three firms already publicly reporting over $1 billion in combined AI-driven value, and the gap between leaders and laggards widening fast.
· 2 min read
Microsoft Copilot Had a Flaw That Stole 2FA Codes
A patched security flaw in Microsoft 365 Copilot, the AI assistant used by tens of millions of enterprise workers, showed how a single email from a stranger could silently pull sensitive data including two-factor authentication codes from a user's inbox without any click or action required.
· 2 min read
Sakana AI Launches 8-Hour Research Agent for Enterprises
Tokyo-based Sakana AI launched its first commercial product this week, an autonomous research agent that runs for up to eight hours and produces reports of up to 100 pages with presentation slides, targeting corporate strategy, finance, and consulting teams.
· 2 min read
SpaceX Buys AI Coding Tool Cursor for $60 Billion
SpaceX has formally agreed to acquire Cursor, a popular AI coding assistant used by major corporations globally, for $60 billion in stock, just days after the company's record IPO, as it tries to rebuild its AI division after a year of serious setbacks.
· 2 min read
EU Requires AI Content Labels on All Products by August
From August 2, 2026, any business operating in the EU that uses AI to interact with customers or publish content must label that content as AI-generated, with fines up to 15 million euros or 3% of global revenue for those that do not comply.
· 2 min read
Databricks Launches Genie One, an AI Data Assistant for Business Teams
Databricks has launched Genie One, a plain-English data assistant that lets non-technical staff ask questions about company data directly in Slack, Microsoft Teams, or a mobile app, without needing to involve IT or data analysts.
· 2 min read
Databricks Launches AI Spending Controls for Enterprise Agents
Databricks used its annual Data + AI Summit this week to expand Unity AI Gateway, a product that gives organizations a single place to see, limit, and audit what their AI agents are doing and spending across multiple AI providers.
· 2 min read
Probably AI Raises $9M to Stop AI Errors at the Source
A new startup called Probably, backed by Andreessen Horowitz, has built a system that catches AI errors before they reach users, and can run on a regular office computer instead of expensive cloud servers, which matters because enterprise AI bills are already spiraling out of control.
· 2 min read
Databricks Launches CustomerLake, an AI-Powered CDP
Databricks has built a customer marketing platform directly inside its data infrastructure, letting large companies run AI-driven campaigns without copying sensitive customer data into yet another separate tool.
· 2 min read
Data Centers Try to Use Less Power to Get Built Faster
A startup called Emerald AI, backed by Nvidia, is testing software that lets data centers voluntarily reduce their electricity use during peak demand, a move that could help new facilities connect to the grid years faster, while a wave of public backlash and new legislation is making the old approach of just consuming as much power as needed increasingly untenable.
· 3 min read
AI Costs Are Now a Budget Problem for Every Company
Companies that gave employees open-ended access to AI tools are now hitting unexpectedly large bills, and the situation is only getting more expensive as AI moves from answering questions to taking multi-step actions on its own.
· 3 min read
Malaysia's Respond.io Raises $62.5M to Enter Western Markets
A profitable Malaysian startup that helps businesses manage customer conversations across WhatsApp, TikTok, and Instagram just raised $62.5 million to expand into North America and Europe, where the shift to messaging-first sales is finally accelerating.
· 2 min read
DOJ Uses National Security to Shield xAI From Air Pollution Suit
The US Justice Department sided with Elon Musk's xAI in a lawsuit over unpermitted gas turbines, arguing that shutting them down threatens military operations, a move that sets a striking precedent for how AI companies could sidestep environmental rules.
· 2 min read
Over 21 Million Songs Were Taken to Train AI
The Atlantic published searchable databases showing more than 21 million songs were fed into AI music tools without artist consent, and the legal fights that followed are now reshaping how the entire music industry handles copyright.
· 2 min read
Google's Gemma 4 Models Now Available on Amazon Bedrock
Amazon Bedrock now offers Google DeepMind's Gemma 4 model family, giving businesses a way to run capable open-weight AI on AWS infrastructure without managing servers or sending data to third parties.
· 2 min read
Facebook Adds AI Search Pulling from Public Posts
Meta has launched AI Mode on Facebook, a search feature that turns public posts and group discussions into direct answers, as the company moves to make its platforms more useful while building toward paid subscription tiers.
· 2 min read
Anthropic Sued Over $200/Month Claude Usage Limits
A paying customer has filed a federal class-action lawsuit against Anthropic, claiming the company's premium Claude Max plans charged up to $200 a month for usage levels that were far lower than advertised, and nearly impossible for subscribers to track in real time.
· 2 min read