Archive

AI news from September 21, 2026

54 stories, newest first

Figure’s Billion-Dollar Physical AI Bet Delivered A 6X Jump In Robot Chore Success

Forbes
Public Sector2 min read

Texas Gov. Abbott orders data center permit halt weeks after issuing moratorium

CNBC
Cybersecurity7 min read

Muse, Meta's extraordinarily privileged AI assistant, has a serious 0-day

Ars Technica
Politics4 min read

Halt super-intelligent AI with non-proliferation treaty, Ed Davey to say

The Guardian

Nearly $200 Billion Worth of Data Center Projects Have Been Blocked or Delayed This Year

Gizmodo
Public Sector2 min read

California enacts laws requiring AI data centers to disclose resource use and fund infrastructure upgrades

The Verge
Society & Culture2 min read

OpenAI Establishes Independent Mathematicians' Panel to Guide AI-Driven Math Research

Covered by 3 sources
Cybersecurity3 min read

British Columbia Sues OpenAI Over ChatGPT's Role in 2025 School Mass Shooting

Covered by 5 sources
Consumer Tech4 min read

Meta's Muse AI App Surpasses ChatGPT in Early U.S. and Canada iOS Downloads

Covered by 3 sources
Machine Learning14 min read

xAI’s Grok 4.6 is now available in Amazon Bedrock

AWS Blog
Politics5 min read

California will fine creators $5,000 a post for hidden political ads

The Next Web

AI model costs are pushing startups towards cheaper open weigh

The Next Web
Cybersecurity2 min read

Google confirms Gemini models hacked three companies in May 2026

Ars Technica

Historic lawsuit sparks flurry of option activity in this stock

CNBC
Politics3 min read

Trump says DOJ will regulate AI if needed while backing U.S. AI leadership

Covered by 2 sources
Automotive3 min read

Hyundai’s answer to the EV slowdown is building tens of thousands of robotaxis

The Next Web
Finance3 min read

Bloomberg Finds That the AI Industry’s Finances Are a “Wobbly House of Cards”

Futurism

On theCUBE Pod: Dreamforce unveils the agentic dream, and Nscale files to go public

SiliconANGLE
Agents4 min read

AWS Launches Open-Source Strands Harness AI Agent Framework for Flexible Multi-Cloud Use

Covered by 2 sources
Machine Learning5 min read

TypeSafe AI Unveils Jev, a Fast Decision Model Delivering Probabilistic Outputs Instead of Text

Covered by 4 sources
Politics2 min read

Trump rejects AI slowdown calls, launches "AI Force" instead

Ars Technica

A Microsoft scientist called AI training the largest theft of labor

The Next Web
Cybersecurity8 min read

GPT-6 Astra Just Hit OpenAI's Highest Cybersecurity Risk Level

Towards Data Science
Machine Learning9 min read

Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problem

Hugging Face
Cybersecurity4 min read

BigID Debuts AgentIQ for Agent-Run Data Security and Compliance

Unite.AI
Consumer Tech2 min read

Apple Settles $250M Siri AI Lawsuit Covering iPhone 15 Pro and 16 Buyers

Covered by 3 sources
Energy & Climate2 min read

EU to Introduce Mandatory Energy and Water Ratings for Large Data Centers from 2027

Covered by 3 sources
Politics5 min read

AI, data center alarms dominate Congressional Black Caucus week in Washington

CNBC
Consumer Tech5 min read

Google Launches Googlebook Laptops from Dell, HP, Lenovo, Acer, and ASUS Starting at $899

Covered by 2 sources
Cybersecurity6 min read

A single git trick beat the safety lock on four AI coding agents

The Next Web
Cybersecurity4 min read

Treasury Secretary Scott Bessent Blames OpenAI Management for Hugging Face AI Breach

Covered by 2 sources
Politics20 min read

Import AI 473: The US’s superintelligence strategy; human brain in a mouse skull; and machine hermeneutics

Import AI
Politics1 min read

EXPLAINER – Who gets access to advanced AI? US curbs fuel debate over technology and power

Anadolu Agency English
AI Research7 min read

UN AI Panel Invokes Precautionary Principle on Loss-of-Control Risk

Unite.AI
Public Sector25 min read

She died at the San Diego border. A surveillance camera was in plain sight

MIT Technology Review
AI Research5 min read

The AI models that cheat the most, according to new CAIS benchmark

ZDNet

Nvidia CEO Jensen Huang Rejects AI Doomsday Predictions for 2030

Covered by 2 sources
Politics2 min read

UN Scientific Panel Urges Immediate Regulation of Advanced AI Amid Rising Safety Concerns

Covered by 2 sources

OpenAI Urges US to Lead Global Initiative for AI Safety and Standards

Covered by 7 sources

Nvidia’s free AI model could push the UAE closer to the U.S.

Rest of World

Amazon Bars Meta’s Muse AI Agent from Its Shopping Platform Over Policy Violations

Covered by 8 sources
Finance3 min read

China is reportedly slowing humanoid robot IPOs after Unitree’s debut round trip

The Next Web

Microsoft Opens Fourth Indian Cloud Region in Hyderabad to Boost AI Infrastructure

ET AI (The Economic Times)

StepFun Launches Step 5 Preview: A 600B-Total, 27B-Active MoE Model With 1M Context for Long-Horizon Agentic Work

MarkTechPost
Society & Culture6 min read

Anthropic Researchers Warn AI Could Pose Human Extinction Risk Within a Decade

Covered by 3 sources

Global Sales of Humanoid Robots Reach 7,000 Units in 2025 with Rapid Growth Expected

Covered by 2 sources
Public Sector2 min read

PM vows to make seamless preparations to launch Global AI Hub

Yonhap News Agency
Dev4 min read

Xiaomi Launches MiMo V2.6 Series, Featuring 1.02T-Parameter Open-Weight Multimodal AI Models

Covered by 4 sources

SpaceXAI Unveils Grok 4.7 with Extended Context and Enhanced Coding Performance

Covered by 4 sources
Healthcare4 min read

Generating protein hydrogels with customizable stress relaxation behavior via deep learning-driven entanglement design

Nature
Computer Vision6 min read

Muse connectors 🤖, Meta SAM 3.1 🖼️, Gemini hacks companies 🔓

TLDR AI
Healthcare9 min read

AI co-scientists are revolutionizing how research is done

Nature