Archive — page 52 of 60
1,797 stories
Murati deposition exposes the 78 texts behind Sam Altman's 2023 ouster
Trial exhibits in Musk v. Altman show the former OpenAI CTO helped trigger the firing, then spent 14 hours texting to reverse it.
OpenAI adds Trusted Contact alerts to ChatGPT amid self-harm lawsuits
Adult users can now designate a friend or family member who gets pinged when ChatGPT detects signs of self-harm in a conversation.
Voi founders raise $16M from a16z for enterprise AI startup Pit
Pit, led by Voi CEO Fredrik Hjelm and ex-CEO Adam Jafer, is building custom back-office automation for industrial clients in Europe.
Thousands of Vibe-Coded Apps Leak Medical Records and Corporate Strategy Decks
Red Access scanned 380,000 apps built on Replit, Lovable and Base44, finding roughly 2,000 leaking sensitive data with no authentication.
Anthropic Teaches Claude Code to 'Dream' Between Sessions
A new idle-time memory consolidation feature lets Claude Code review up to 100 prior sessions and lifts task success by up to 10 points.
Google Shuts Project Mariner, Folds Browser Agent Into Gemini
Google has quietly retired its agentic browser project, citing the steep compute cost of screenshot-based automation versus code-driven agents.
Red Access Finds 2,000 Vibe-Coded Apps Leaking Medical and Corporate Data
Israeli security firm scanned 380,000 apps built on Lovable, Replit and Base44 and found roughly 2,000 with no authentication at all.
Musk lawsuit puts OpenAI safety record on trial as ex-staffer testifies
Former AGI readiness lead Rosie Campbell told a federal court OpenAI shifted from research to product, with the Super Alignment team shut down in 2024.
Perplexity opens Personal Computer to all Mac users with 400 connectors
The local AI agent app exits its Max-tier waitlist, runs on-device, and can be triggered remotely from an iPhone.
NVIDIA and DOE stake Genesis Mission on a 100,000-GPU Solstice supercomputer
Argonne's Solstice will hit 5,000 exaflops on Vera Rubin GPUs — five times the entire TOP500 list combined — to run AI for science.
Aurora moves from a handful of driverless trucks to hundreds in 2026
CEO Chris Urmson says the autonomy case cracks on long-haul freight before robotaxis, with commercial operations now a year old on the Dallas–Houston lane.
Spotify opens its library to AI-generated personal podcasts via a new CLI
A beta command-line tool lets Codex, Claude Code and OpenClaw users pipe agent-made audio briefings straight into their Spotify library.
Moonshot AI raises $2B at $20B valuation as Kimi climbs OpenRouter charts
The Beijing lab's valuation has doubled to $20B in months as its Kimi K2.6 model becomes the second-most-used LLM on OpenRouter.
Google launches $100 screenless Fitbit Air and Gemini-powered Health app
First Fitbit hardware in four years arrives with a rebranded app, a $10/month AI coach, and the death of the Fitbit name across Google's wearable stack.
Stick Figure's 'Angels Above Me' goes No. 1 — but AI remixes are stealing the royalties
A seven-year-old reggae track topped iTunes in six countries this week. The band says it isn't getting paid for any of the four viral AI versions.
RedAccess finds 5,000 vibe-coded apps from Lovable, Replit and Base44 leaking data
Roughly 2,000 of the AI-built apps exposed medical records, customer chat logs and corporate strategy decks to anyone with the URL.
Musk tried to move OpenAI's founders into Tesla, 2018 emails show
Court testimony reveals Musk pitched Altman, Brockman and Sutskever on running AI inside Tesla — or making OpenAI a Tesla subsidiary.
Spotify's AI DJ adds French, German, Italian and Brazilian Portuguese
The interactive feature jumps from 2 to 6 languages and lands in 8 new markets, reaching more than 75 countries.
ASML, Google Cloud and Applied Intuition CEOs warn AI is supply-limited for years
ASML's Christophe Fouquet says chips will be supply-constrained for two to five years as Google Cloud's backlog jumps from $250B to $460B in a quarter.
CAISI signs Google, Microsoft and xAI to AI safety tests after Mythos scare
Trump's renamed AI safety body locks in pre-deployment evaluations with three frontier labs as a mandatory testing executive order looms.
TSMC signs 30-year deal for all of Hai Long offshore wind output as AI chip demand strains Taiwan grid
The chipmaker is locking in 1 gigawatt of offshore wind as its share of Taiwan's electricity heads toward 25% by 2030.
Anthropic taps SpaceX's 220,000-GPU Colossus to lift Claude Code limits
Deal gives Anthropic 300 megawatts of new capacity from SpaceX's Memphis data center, enough to double Claude Code windows for Pro and Max users.
OpenAI's Compute Bill Hits $50 Billion as Brockman's Stake Reaches $30 Billion
Greg Brockman testified under oath that OpenAI will spend $50 billion on compute in 2026, up from $30 million in 2017.
Brockman Tells Court OpenAI Will Spend $50B on Compute in 2026
OpenAI president's sworn testimony reveals a compute bill that has ballooned from $30 million in 2017 to $50 billion this year.
OpenAI Opens Self-Serve Ads Platform With $2.5B Revenue Target
OpenAI's ads manager goes live for all US advertisers at $3 to $5 per click, with internal goals of $100 billion in ad revenue by 2030.
OpenAI Opens Self-Serve Ads Platform, Targets $100B by 2030
OpenAI scraps its $50,000 minimum spend and opens ads.openai.com to all US advertisers, chasing YouTube-scale revenue.
OpenAI's GPT-5.5 Instant Cuts Hallucinations 52% as Default ChatGPT Model
OpenAI replaces GPT-5.3 Instant with a new default model touting sharper accuracy in medicine, law, and finance, plus user-visible memory controls.
Chrome Caught Silently Pushing 4GB Gemini Nano Model to User Devices
An EU researcher says Google installed a Gemini Nano model on Chrome users' machines without consent, raising GDPR and ePrivacy questions.
Apple to Open Siri to Claude, Gemini and ChatGPT at WWDC as $100-a-Device Settlement Looms
iOS 27 will let users route Siri queries to third-party models, while Apple faces a payout to iPhone 15 Pro and 16 buyers over Apple Intelligence delays.
Chrome Quietly Pushed a 4GB Gemini Nano Model to User Devices, Researcher Finds
An EU-based security researcher says Google installed a four-gigabyte AI model on Chrome users' machines without a consent prompt.
Stay ahead of everyone in AI.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at