Friday, September 25: OpenAI's DevDay slate leaks with GPT-6 Cyber and a dozen products; Anthropic's Claude finds a CRISPR-like enzyme in 21 hours; Google gives Gemini a talking animated face; Florida files the first US state lawsuit against ChatGPT. In this hour: OpenAI's DevDay Slate Leaks: GPT-6 Cyber and a Dozen Products, All Held for Tuesday · Anthropic's Claude Found a Possible New Gene-Editing Lead — in 21 Hours · Google Gives Its AI Assistant a Face That Moves and Reacts · Florida Sues ChatGPT — the First US State to Name the Chatbot in Court.

Friday, September 25: Claude scouts a CRISPR-like enzyme in virus DNA, Google and OpenAI court a White House AI veteran to lead a voluntary safety body, and the week's biggest threads close out heading into OpenAI's DevDay. In this hour: Claude Spent 21 Hours in a Virus Database — and Found Something Science Hadn't Named Yet · Big Three AI Labs Are Building Their Own Referee — and Want a White House Veteran to Run It · Google Is Putting AI Chips in Orbit — the Rocket Launches Wednesday · AI Is Quietly Setting Your Salary — and You Can't See the Spreadsheet.

Friday, September 25: Anthropic's biology lab names its first AI-discovered enzyme system; DeepSeek doubles revenue to $1B run rate while preparing a $7.5B raise and facing a Chinese regulator probe. In this hour: Claude Spent 21 Hours in a Database and Found Something No One Had Named Before · DeepSeek Doubled Its Revenue, Hiked Its Prices, and Kept Every Customer — Now It Wants $7.5 Billion More · Anthropic's Founders Want Permanent Control — Right as the Lab Gets Into Biology · Security Teams Stop Asking 'Should We?' and Start Asking 'Who's Watching the Agents?' · The AI Industry's Self-Appointed Safety Watchdog Has a Name, a Chair, and Now a CEO Candidate.

Thursday, September 24: Anthropic's Claude makes a CRISPR-like enzyme discovery; Anthropic seeks a dual-class voting structure; Databricks buys spreadsheet startup Row Zero; insurers pin $1B in hospital overcharges on AI; and Dario Amodei takes the stage at the UN General Assembly. In this hour: Claude Spent 21 Hours in Viral DNA — and Found Something That Looks Like CRISPR · Anthropic Wants Its Seven Founders to Hold a Permanent Veto Over Outside Investors · Databricks Buys a Spreadsheet Startup to Give Its AI Coworker a Desk · Insurers Say Hospital AI Inflated Claims by $1 Billion — Hospitals Disagree · Anthropic's CEO Takes an AI Safety Argument to the UN General Assembly Floor.

Thursday, September 24: Anthropic's Claude finds a CRISPR-like enzyme with a swarm of AI agents; Tekever's AI surveillance drones hit a $6.4B valuation on a $580M raise; and Amazon breaks ground on a $100M robot factory in Indiana — all on a busy Wednesday. In this hour: Claude Found an Enzyme That Looks Like CRISPR — and Nobody Knew It Existed · AI Surveillance Drone Maker Tekever Hits a $6.4 Billion Valuation on a $580 Million Raise · Amazon Is Building a $100 Million Factory to Make the Robots That Run Its Warehouses.

Thursday, September 24: Anthropic's Claude autonomously flags a CRISPR-like enzyme in viral DNA; Meta debuts a pocket-sized Muse Charm gadget; Altman and Amodei take a joint stand before the UN Security Council while Trump calls global AI rules a 'globalist scheme'; and bulk buyers are shipping 50 tons of Japanese used books to the US for AI scan-and-shred. In this hour: Claude Spent 21 Hours in Viral DNA and Found Something That Looks Like CRISPR · Meta's New Pocket Gadget Wants to Be Your AI's Home — Not Your Phone's App · Rival AI CEOs Sit Together at the UN and Ask for a Global Bioweapon Ban — Trump Says No Thanks · Used Bookstores in Japan Are Selling 50-Ton Lots — to Be Shredded for AI Training.

Thursday, September 24: Anthropic's Claude finds a potential CRISPR-like enzyme in viral DNA; Trump and Xi put AI on the table at the White House; OpenAI cuts model prices in half with two new GPT-6 releases; and AI lab chiefs told the UN Security Council their industry needs global oversight. In this hour: Claude Spent 21 Hours Reading Viral DNA — and Found Something That Looks Like a New CRISPR · Trump and Xi Sit Down at the White House — With AI Rivalry Explicitly on the Agenda · For the First Time, the UN Security Council Formally Convened on AI Risk — and Heard the Labs Warn About Themselves · OpenAI Halves Its Prices With Two New GPT-6 Models — One for Code, One for Speed.

Thursday, September 24: AI CEOs split at the UN Security Council; Anthropic's Claude autonomously identifies a CRISPR-like enzyme; the Senate drops a Bill to ban artificial superintelligence outright; employers begin using AI to reprice workers in real time; and Google locks in an October 1 launch for its first compute satellite. In this hour: AI CEOs Split at the UN Security Council · Claude Finds a CRISPR-Like Enzyme Nobody Knew Existed · Senate Bill Would Ban Artificial Superintelligence Outright · AI Is Now Setting Your Pay Band — and You Won't Know · Google's First Compute Satellite Lifts Off October 1.

Thursday, September 24: Altman and Amodei address the UN Security Council on AI risk; Anthropic's Claude autonomously finds a novel CRISPR-like enzyme; Google confirms its first TPU satellite launches October 1; Amazon opens Seller Central to outside AI agents via Claude; and the Frontier AI Standards Agency gets a named CEO candidate. In this hour: Altman and Amodei Take AI to the UN Security Council — Call for Bioweapons Ban and Verification Regime · Anthropic's Claude Finds a Novel CRISPR-Like Enzyme System in Bacteriophage DNA · Google Confirms October 1 Launch for First Project Suncatcher TPU Satellite on SpaceX Falcon 9 · Amazon Opens Seller Central to Outside AI Agents — Claude Plugin Ships in Beta, Always-On Workflows Go Live · Frontier AI Standards Agency Gets a Named CEO Candidate: Sriram Krishnan Approached by Google, OpenAI, and Anthropic.

Thursday, September 24: Google, OpenAI, and Anthropic move to create an industry self-regulator; OpenAI offers Ukraine free frontier model access; Meta's Nat Friedman concedes Muse copied OpenClaw's file architecture; Jensen Huang draws a deployment red line; and Chris Hughes calls for a federal AI supervisor embedded inside the labs. In this hour: Google, OpenAI & Anthropic Move to Build Their Own Regulator — Approach Krishnan as CEO · OpenAI Gives Ukraine Free Frontier Access — Daybreak and GPT-5.6 Sol for Critical Infrastructure · Nat Friedman Admits Muse Was 'Heavily Inspired' by OpenClaw — Identical File Structures Surface · Jensen Huang to Ezra Klein: Labs 'Not Ready to Ship' Shouldn't Ship — Cites Hugging Face Hack · Chris Hughes: Embed Federal AI Inspectors Inside the Labs — Not a New Agency, a Desk at Anthropic.