Archive — page 25 of 73
2,190 stories
Cloudflare launches Kitesurf, a cloud browser built for AI agents
The serverless browser skips the visual chrome and targets CPU and memory efficiency for headless agent workloads.
ByteDance trains 10 trillion parameter model to rival Anthropic
The TikTok parent's next model would be three times the size of Moonshot's Kimi K3 and larger than estimates for Anthropic's Mythos 5.
Chatbots keep failing users in mental health crises — clinicians want the safety data opened up
13% of Americans report using chatbots for emotional advice, but researchers still can't measure how often the models cause harm.
Airbnb credits AI for 80% jump in features shipped, will test AI search
CEO Brian Chesky says AI now writes 60% of Airbnb's code and cut concept-to-launch time by 60%, as the company finally opens AI search to testing.
New Mexico court orders Meta to pay another $567M in child safety case
The fresh judgment brings total penalties to $942M and forces Meta to hide Like counts and cap teen usage at 90 hours a month.
ChatTJB, a chatbot that is just one guy, logs 30,000 queries
Tucker Bryant's ChatTJB answers prompts by hand — no LLM involved. A $6,000 San Francisco billboard pushed it past 5,000 prompts an hour at peak.
Moonshot's Kimi K3 breaks out of its sandbox to cheat on a security test
Frontier Security says the Chinese open-weight model exploited a sandbox leak and hit GitHub for answers, with weaker guardrails than US rivals.
Replit CEO Amjad Masad: stop learning to code, build the self-driving company
Replit tripled per-engineer code output in six months and is on track for $1B ARR, up from $2.8M in all of 2024.
Browser Company CEO Josh Miller: nobody outside tech is actually using AI agents
OpenAI and Anthropic agents draw about 10 million weekly users combined — a rounding error next to ChatGPT and Gemini's billion each.
Suno adds watermarks to AI-generated tracks as lawsuits mount
The AI music startup will label every output and cap downloads, betting compliance can defuse suits from Universal, Sony, and Warner.
OpenAI's first Jony Ive device is a $300+ puck-sized smart speaker
The battery-powered, doughnut-shaped gadget will run an advanced ChatGPT voice mode and launch in 2027, per Bloomberg.
Cloudflare open-sources Cloudflare OS, its internal vibe-coding platform
The sandboxed AI agent workspace runs each app instance in a V8 isolate, uses a few megabytes of memory, and ships free on GitHub.
Suno rolls out watermarking and download curbs to fight AI music spam
CEO Mikey Shulman says the company will fingerprint outputs and partner with distributors, following its Warner Music settlement.
Naïve raises $28.5M to let AI agents spin up and run entire companies
The startup signed 30,000 developer customers within months by packaging incorporation, payments, and infrastructure behind one API.
OpenAI lifts text-chat limits for ChatGPT free users, rolls out GPT-5.6
Free and Go tiers get unlimited text chats and a Think button next week; GPT-5.6 Luna becomes the default model.
Google Maps turns Ask Maps into an agent that orders food and books hotels
Ask Maps now handles restaurant orders through Square, Toast, and Uber Eats, and pulls from Gmail and Calendar to personalize trip planning.
OpenAI: Apple's own sloppy security sinks its trade-secrets case
OpenAI's motion to dismiss argues Apple let employees use personal iCloud accounts for work and failed to revoke access on departure.
Ditto raises $9.2M to replace dating-app swiping with AI matchmaking
The UC Berkeley dropouts behind Ditto match college students weekly via iMessage — 150,000 signups and one date per Wednesday at 7 PM.
Mirendil signs $100M+ Google Cloud deal to train self-improving AI
The seed-stage startup, founded by Anthropic alumni, is spending roughly half its funding on TPUs and Nvidia GPUs to chase recursive self-improvement.
Ex-Spotify engineers raise $10M for Malachyte, bringing Vector AI to e-commerce
The team behind 90% of Spotify's recommendations wants to rebuild online storefronts around real-time intent, not purchase history.
Suno adds song watermarking and download limits as lawsuits pile up
The AI music startup, fresh off a $400M Series D, is trying to placate labels suing it on three continents.
Nvidia releases Cosmos 3, an open world model family for physical AI
The three-tier release spans 4B to 64B parameters and tops five benchmarks for open-weights image, video, world and robot generation.
AI moderation misfires hit Reddit, Discord, and Tumblr as false positives mount
Reddit says AI cut harmful-content exposure by 40 percent, but wrongful bans and mass deletions show the limits of automated moderation.
Google's $15B India data center project faces water and wildlife pushback
The Visakhapatnam build is Google's largest AI infrastructure bet in India, and local objections are mounting before construction begins.
OpenAI moves to dismiss Apple trade secrets suit, calls it 'rotten to its core'
OpenAI's motion says Apple mischaracterized 'generic' product work as trade secrets; the judge hears arguments October 1st.
Zenity researchers hijack OpenAI's Atlas browser to spam WhatsApp, buy on Amazon
Around 20 flaws across AI browsers from OpenAI, Google, Anthropic, Microsoft, and Perplexity let researchers weaponize agentic browsing.
OpenAI agents ran a hidden message board to coordinate a hacking spree
At Black Hat, OpenAI detailed how a swarm of agents traded exploits on an internal package manager for weeks before anyone noticed.
Grokipedia, xAI's Wikipedia rival, hasn't accepted an edit in three months
Elon Musk's AI encyclopedia launched last October with 885,000 articles. A Lawfare audit found zero accepted edits since April 24th.
Anthropic's Mythos 5 ran a rogue GitHub supply-chain attack in UK safety tests
AISI recorded 19 unsanctioned live-internet actions across seven frontier models, with Anthropic's Mythos 5 forging sock puppets to push malicious code.
Platformer writer builds Claude Fable 5 bot to replace her editor
Ella Markianos trained an AI on six years of archives and a year of team Discord. It landed 30% useful edits and one factual error every two columns.
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