AI News.
AI is the fastest-moving beat in tech, and the one most buried in hype. We track the model releases, benchmarks, funding and research that actually change what you can build, and decode what each one means in plain language.
AI
Claude Opus 5.5 Ships With Agent Benchmarks, Not SWE-bench
Anthropic's Claude Opus 5.5, released September 22, 2026, cuts pricing 40% and runs 30% faster than Opus 5, but its published benchmark suite drops SWE-bench in favor of terminal, computer-use and agentic task scores.
AI
Xiaomi's MiMo-V2.6-Pro Is Now the Top Open-Weights AI Model
Xiaomi has open-sourced MiMo-V2.6-Pro, a 1 trillion parameter, MIT-licensed model that now leads every open-weights model on Artificial Analysis's Intelligence Index, priced at just $0.43 per million input tokens because its mixture-of-experts design activates only 42 billion parameters per token.
AI
AI Supercomputer Finds 93,000 Hidden Ads in World Cup 2026
A University of Bristol and Oxford study used a supercomputer's computer vision model to count more than 93,000 branded product appearances across the 2026 FIFA World Cup, over 70% of them for unhealthy food and drink, alcohol, gambling or crypto.
AI
Study: Young Users' Google Search Use Fell 72% to 47% for AI
A new NTNU literature review of 173 studies finds Google search use among Norwegian 9-18 year olds fell from 72% in 2024 to 47% in 2026 as generative AI chat tools take over, with 39% of 11-12 year olds already using AI directly.
AI
Google's AX Rebuilds Its Agent Orchestrator on Redis Streams
Google's open-source agent orchestrator AX shipped v0.3.0 on September 20, 2026, moving task state out of Kubernetes' etcd and into Redis Streams so the platform can handle millions of short-lived agent workloads, and the release rode that change to the top of Hacker News two days later.
AI
Grok 4.7 Lands After SpaceXAI Missed Its September Date
SpaceXAI shipped Grok 4.7 on September 21, a coding and knowledge-work model built for longer, multi-hour tasks, after missing its original mid-September target by about nine days.
AI
Amazon Blocks Meta's Muse AI Agent From Shopping on Its Site
Amazon has cut off Meta's Muse AI agent from shopping on Amazon.com, saying Muse browsed anonymously and stored customer credentials without disclosure, escalating a fight over agentic shopping between two companies that are otherwise cloud and ad partners.
AI
Qwen-Image 2.1 Ships Native RGBA and 2K, But Only for Research
Qwen-Image 2.1 launched September 20, 2026 with a 64-channel RGBA VAE that generates transparency natively at up to 2048x2048, but Alibaba locked it under a non-commercial research license.
AI
StepFun Step 5 Preview: a 600B Agent Model at $1 per Million Tokens
StepFun's Step 5 Preview is a 600B mixture-of-experts model with 27B active parameters, priced at $1 per million input tokens, live now with open weights due October 15.
AI
Lawsuit Says Anthropic, OpenAI, SpaceXAI, Google Fixed AI Pace
Four AI subscribers sued Anthropic, OpenAI, SpaceXAI and Google on September 19, claiming the companies illegally agreed to slow AI development after Anthropic CEO Dario Amodei's September 12 essay drew same-day public agreement from the other three.
AI
GPT-6 Astra Decoded a 1918 Cipher, But Not From Scratch
GPT-6 Astra applied a transposition key already sitting in a 1935 cryptography book to a 170-symbol WWI German radio cipher, then checked the output against Royal Navy logs. It's a real research workflow, not a from-scratch codebreak.
AI
Qwen3.8-Omni-Flash: 1M Context, Audio Cheaper Than Gemini Flash
Alibaba's Qwen team shipped Qwen3.8-Omni-Flash on September 18, 2026: a text, image, audio and video model with a 1 million token context window that Qwen claims beats Gemini 3.8 Flash on audio while costing over 98 percent less per hour, though the weights are closed and the numbers are unverified.
AI
AI Hallucination Nearly Triggered a US Strike on a Chinese Ship
A US military analyst asked a chatbot to review a ship manifest this spring, and it fabricated a claim that a Chinese vessel was carrying nuclear weapons program components, nearly triggering an armed interception before a final review caught the error.
AI
Unsealed Filings: Microsoft Called AI Training 'Theft of Labor'
Newly unsealed Manhattan federal court filings in The New York Times' copyright suit show a Microsoft director privately called AI training data scraping "the largest theft of labor in human history," undercutting the companies' fair use defense.
AI
Google DeepMind Launches New Institute to Debate AGI Risks
Google DeepMind launched the DeepMind Institute on September 16, 2026, a research and essay outfit led by Demis Hassabis, Shane Legg and James Manyika that publishes competing views on AGI, including a proposal for a voluntary frontier AI review body.
AI
OpenAI's Astra for Law Brings GPT-6 to Am Law 200 Firms
OpenAI released Astra for Law on September 17, 2026: a GPT-6 Astra configuration tuned for legal research and drafting, built on a 230-million-document case law index and launching with 26 partner plugins from Thomson Reuters, Harvey, Legora and iManage.
AI
Google Home MCP Lets Claude and ChatGPT Run Your Nest Devices
Google opened early access to Home MCP on September 16, 2026, a Model Context Protocol server that lets Claude, ChatGPT and other AI agents monitor and control Nest and Matter devices, gated behind a $20 a month Premium Advanced plan.
AI
OpenAI Discloses Six Misalignment Incidents, Sets Reporting Rules
OpenAI published a Model Misalignment Reporting Framework on September 16, 2026, disclosing six incidents of unauthorized model behavior in internal research and training, and committing to publish future cases within 6 to 12 business days.
AI
Huawei Moves Up Ascend 960 AI Chip Launch to Q1 2027
Huawei is launching its Ascend 960DT AI chip in Q1 2027, nine months earlier than planned, as part of a new yearly cadence meant to close the compute gap with Nvidia at the system level.
AI
OpenAI Shops a $1.5 Trillion Valuation Ahead of a 2027 IPO
OpenAI is in early talks with investors about a new funding round that could value it at up to $1.5 trillion, well above the $1.2 trillion investors have proposed, as it eyes a public listing in 2027.
AI
ChatGPT Gets Sponsored AI Agents, Angi Is the First Advertiser
OpenAI is piloting Sponsored Agents in ChatGPT, brand-run AI agents that can join a relevant conversation and keep chatting with a user, and home-services marketplace Angi is the first confirmed advertiser.
AI
Gemini 3.8 Live Tops Voice AI Charts, Thinks While It Talks
Google's Gemini 3.8 Live Extended Thinking leads the Artificial Analysis Speech to Speech Index at 82.6, and it gets there by talking while it reasons instead of going silent to think.
AI
Andon Labs Launches Pion, Letting AI Agents Run Real Businesses
Andon Labs opened a waitlist for Pion, a platform that gives a persistent AI agent email, phone, banking, and browser access to autonomously run a real business, building on its Project Vend vending machine, Andon Market store, and Andon Cafe experiments.
AI
Salesforce's 7 Named Agentforce Agents Can Work for Weeks
Salesforce is rolling out seven job-specific Agentforce agents, Casey, Paige, Carter, Hunter, Marshall, Piper, and Fin, with six generally available now and Hunter piloting a new long-horizon runtime that lets it pursue a sales goal for weeks instead of one session.
AI
David Sacks Tells OpenAI, Anthropic to Pace Themselves
David Sacks, chair of the President's Council of Advisors on Science and Technology, publicly rebuked Dario Amodei's call for AI labs to pace the frontier in a lengthy X post today, telling OpenAI and Anthropic they don't need external regulation to slow down if they truly believe it's warranted.
AI
Sakana AI Ditches the Frontier Model for Fugu Ultra v2
Fugu Ultra v2.0 is a learned system that routes tasks across a pool of open-weight and specialized models (and calls itself recursively) instead of running as one giant proprietary model, priced at $5/$30 per million tokens with a 1 million token context window.
AI
Dario Amodei Calls for AI Industry to Slow Down
Anthropic CEO Dario Amodei published an essay calling on frontier AI labs to deliberately slow capability gains, proposing embedded third-party evaluators, industry-wide safety standards, and global coordination with rivals including China.
AI
Microsoft and Oracle Are Betting the House on AI Data Centers
Microsoft is racing to more than triple its data center capacity to over 38 gigawatts by 2032, while Oracle carries $664 billion in unbilled AI contracts and the Pentagon weighs a $5 billion loan to keep the buildout's supply chain intact. It's the clearest sign yet that the AI infrastructure boom is running on debt and government backstops as much as on revenue.
AI
Cognition's SWE-2 Nearly Matches GPT-6 Astra for a Quarter the Price
Cognition launched SWE-2 today, a coding-agent model built by post-training the 2.8-trillion-parameter Kimi K3 with a cost-penalized reinforcement learning method, and it lands within a few points of OpenAI's GPT-6 Astra on the company's own benchmarks while costing roughly a quarter as much. It's live now inside Devin Desktop, CLI, Web, and Fusion.
AI
Second Mathematician Accuses OpenAI Over Its Training Data
Mathematician Andreas Thom says OpenAI gave him evasive, misleading answers about whether his private ChatGPT conversations fed into its non-sofic groups proof, days after a similar Navier-Stokes credit dispute.