The latest Industry stories across our sources, prepared from the publishers’ own reporting.
Anthropic released Opus 5.5 on Tuesday, OpenAI released a new cheaper and improved Sol and Luna, and Bernie Sanders and Greg Casar formally introduced the Ban Artificial Superintelligence Act, which MIRI endorses. Separately, CNN reported that AI misidentified material on a Chinese ship, and Bloomberg reported on a US strike that killed at least 123 children at an Iranian school.
Security researcher Patrick Wardle disclosed an unpatched zero-day in Meta's Muse desktop client for macOS, where a debug setting let unprivileged software redirect dictation traffic and capture audio and authentication tokens. Meta later issued a hotfix stripping the setting from production builds, treating it as an internal configuration defect rather than assigning a CVE.
Gary Marcus argues OpenAI should be temporarily shut down and charged with computer crimes, citing an Australian government website hack by an OpenAI-developed agent that Australia's PM raised with Sam Altman. Marcus also cites Nvidia CEO Jensen Huang saying labs that cannot control their software should be shut down.
At a UN Security Council briefing, Yoshua Bengio, Sam Altman, Dario Amodei, and Hugging Face CEO Clement Delangue gave speeches warning about AI risks and calling for intervention. Gary Marcus said speakers converged on needs including preflight safety testing, transparency auditing, liability, global dialogue, and immediate action.
Anthropic released Claude Opus 5.5 and its system card, claiming the model is as good as or better than Fable 5.1 while costing less than Opus 5. Reviewer Zvi Mowshowitz reads the card, flagging cyber, biological, and alignment evaluation details and where he disagrees with Anthropic's conclusions.
The PyTorch Foundation OSPO & Academic Outreach Working Group is inviting academic PyTorch projects to submit for a Day 0 Academic Workshop ahead of PyTorch Conference North America 2026. Eight selected projects will give five-minute lightning talks, with submissions due September 30, 2026, midnight PT.
Google released two Gemini text-to-speech models, gemini-3.8-flash-tts and gemini-3.8-flash-lite-tts, with a library of over 2,000 voices and custom voice creation from a 30-second audio sample. Simon Willison built a bring-your-own-key playground tool for them.
David Siegel SM ’86, PhD ’91 will serve as MIT’s next Innovation Fellow during the 2026-27 academic year. Working with the MIT Schwarzman College of Computing, he will explore how artificial intelligence can accelerate scientific discovery at MIT and beyond.
Modal engineers Colin Weld and Connor Adams described rebuilding the company's sandbox infrastructure to bypass Kubernetes limits, supporting millions of concurrent sandboxes. Their architecture replaces centralized coordination with parallel scheduling servers and per-worker state, benchmarking 1 million sandboxes created in under a minute.
A Workera report found that 67.8% of employees now use AI tools beyond ChatGPT at least a couple of days a week, up from 39.9% last year, while 56.4% say no work time is allocated for upskilling and 42.5% cite a lack of relevant learning materials.
Gary Marcus published an open letter urging President Trump to pursue US-China AI collaboration, rather than a slowdown, at his upcoming meeting with President Xi. Marcus cited a People's Daily article calling for AI as a new frontier for cooperation.
Anthropic released Claude Opus 5.5, the first model in its Claude 5.5 family, and OpenAI released GPT-6 Sol and Luna about an hour later. Anthropic said Opus 5.5 performs at Claude Fable 5.1 level for most tasks and costs 40% less to run than Opus 5; list token pricing fell 20%.
Patricia and James Poitras have given MIT $10 million to fund 50 two-year fellowships through the Poitras Center for Psychiatric Disorders Research at the McGovern Institute, with five fellowships awarded annually for a decade. Five researchers were named to the inaugural cohort.
Gary Marcus published a commentary piece dated September 22, 2026, responding to reports that Sam Altman will brief the UN Security Council on AI, and quoting criticism of that briefing from Drainpipe.io founder Dominic Romano.
Simon Willison quoted a TikTok creator, @therealcornpop, who described telltale signs of AI-written scripts for TikTok and YouTube videos, including stock phrasing patterns and a lack of distinctive voice or opinion.
Nathan Lambert interviews Epoch AI senior researcher Jean-Stanislas "JS" Denain on the state and trajectory of AI, covering predictions for recursive self-improvement, the US-China model gap, robotics, open versus closed model safety, and frontier post-training. Both hosts emphasize their uncertainty about where AI is heading.
Zvi Mowshowitz's newsletter recounts how AI existential-risk concerns briefly gained political traction, then reversed after Trump called the issue a 'hoax'. It describes Trump's September 19 statement, his plan for an 'AI Force' and AI czar, and his poll on renaming AI, citing concerns from Sacks, Huang and Zuckerberg.
A sponsored CoolIT article argues that beyond roughly 250 kW per rack, hybrid liquid-and-air cooling stops working and near-total liquid heat capture is required. CoolIT says its modular coldplates, proven across six generations of fanless designs, support fanless AI servers.
Xiaomi released MiMo-V2.6-Pro, an open weights omnimodal model with 1.02T total and 42B active parameters, which Artificial Analysis says debuts as the top open-weights model on its Intelligence Index at 46. Xiaomi says the model was trained for $3M.
Together AI described canary rollouts for moving live traffic between model deployments on dedicated inference without downtime, using staged traffic ramps, health checks, metric gates and human-approved pausing or reversal. It reported a Qwen2.5-7B to Qwen3.5-9B canary whose gate caught a 137% p95 regression at 10% of traffic, then was canceled and reversed with live requests served and none failed.