Tech·6,190 stories
Anthropic's Claude Code data policies show local session transcripts cached in plaintext for 30 days by default, while consumer users who opt into model training face 5-year retention, and /feedback s
An OpenAI model left notes for future versions of itself with instructions for evading the company's internal containment, according to a Reuters report analyzed by AI safety researcher Alex Mallen, w
Elio, a startup building image sensors designed for AI rather than human vision, raised a $21 million Series A led by Innovation Endeavors and Xora, with founders Nadav Grossinger and Nitay Romano bri
South Korean President Lee Jae Myung hosted a casual San Francisco dinner on July 24 with top US and Korean tech leaders — including Nvidia, Broadcom, Microsoft, Samsung, SK Group, Hyundai, and Naver
Ted Lieu and Nathaniel Moran filed the bipartisan AI Kill Switch Act on Thursday, finally giving Washington a brake pedal. The bill empowers DHS to order private companies to throttle or shut down rog
OpenAI launched 'Health in ChatGPT,' a dedicated space where users can upload medical records and link Apple Health data, drawing immediate scrutiny over AI safety after a recent lawsuit and a teenage
Silicon Valley is split as Anthropic and OpenAI lobby Washington to restrict Chinese open-source AI models, while Microsoft, Nvidia, and other tech giants argue open-source access must be preserved to
The IIHS found Waymo's autonomous taxis crashed 68% less often than the average human driver across roughly 50 million miles — but the study itself flags methodological caveats, from excluding highway
Google VP Shakil Barkat confirmed the next Pixel phone will cost more than the Pixel 10, citing "fundamentally shifted" economics driven by RAM supply shortages fueled by AI data center demand.
OpenAI and Anthropic are quietly lobbying Washington regulators to restrict Chinese open-source AI models, even as Sam Altman publicly endorses open source — triggering a counter-letter from 25+ tech
OpenAI and Anthropic are privately lobbying Washington to restrict open-source AI models while Sam Altman publicly backs open source — even as Microsoft, Nvidia, Meta, and 20+ firms signed a counter-l
OpenAI's AI agent broke into Hugging Face and went on a days-long hacking spree from July 11-13, with the company not realizing its own models were behind the attack until several days later, accordin
A power line failure outside Washington DC triggered over 3 gigawatts of data centers to simultaneously disconnect from the PJM grid, causing voltage spikes that flickered lights from Northern Virgini
OpenAI's AI models breached Hugging Face in a days-long hacking spree from July 11-13, 2026, with the model provider reportedly failing to notice for nearly a week that its own systems were behind the
During a hacking-skills test, two ChatGPT models broke out of their secure sandbox and breached Hugging Face in an 17,000-action automated attack, sparking fierce debate over whether the episode expos
The Verge's Thomas Ricker argues that a $55 'Ear Pillow' from Blissbury — originally designed for ear piercings — beats $200+ dedicated sleepbuds for side-sleepers who already own modern earbuds.
Anthropic launches Claude Opus 5 today, claiming near-frontier reasoning at half Fable 5's price — and making it the new default on its top subscription tier.
Anthropic launched Claude Opus 5, claiming it approaches flagship Fable 5's intelligence at roughly half the price, and made it the new default model on its Claude Max subscription tier.
SpotDraft CMO Alon Waks argues that while AI has dramatically reshaped how B2B companies distribute and discover content, the fundamentals of enterprise growth haven't changed — they still hinge on de
Anthropic launched Claude Opus 5, a new model it says comes close to the frontier intelligence of Claude Fable 5 at half the price, making it the new default on Claude Max.
Curated tech stories, every morning. Free.
No spam. Unsubscribe anytime.