[AI DAILY NEWS RUNDOWN] OpenAI Claims Navier-Stokes Math Proof Amid Theft Accusations, US Calls Out Chinese Model Distillation, & AI Drug Reverses Aging (Sept 09, 2026)

Before we start, I need your vote:

My free bedtime-story app DjamgaMind Kids (600+ ad-free stories on Black history, African & Caribbean folktales) is a Top 10 finalist in Melamoon, a national pitch competition for Black founders in Canada. Next round is a public vote: Top 5 go to the Toronto Grand Finale in October. One vote, ~20 seconds, one per person, closes Sept 22:

Visit https://djamgamind.com/pitch to vote

Profile: https://melamoon.ca/top-10-finalists-vancouver/djamgamind.

Thank you!

#Melamoon2026 #DjamgaMind

🎧 Listen/Watch ADS-FREE: https://podcasts.apple.com/us/channel/djamgamind/id6760446113

Summary: In today’s briefing, we analyze “Scientific Autonomy, Distillation Diplomacy, and the Intellectual Property War.” We deconstruct OpenAI’s controversial Navier-Stokes proof running 10,000 agents and the resulting plagiarism allegations. We examine Washington accusing six Chinese labs of distilling US frontier models. We evaluate Insilico’s AI drug rentosertib reversing biological age markers, Anthropic researchers raising existential risk alarms, Meta’s launch of the “Muse” personal agent, and NBC polling indicating 70% of Americans fear AI’s trajectory.

Important Topics:

• OpenAI Claims Navier-Stokes Solution Amid Controversy: OpenAI reports an internal model running 10,000 agents for 88 hours generated a 100-page Lean-verified proof for the Navier-Stokes equations. NYU’s Tristan Buckmaster accuses OpenAI of using his private Codex drafts to replicate the breakthrough.

• US Accuses Chinese Labs of Systematic Model Distillation: Washington charges DeepSeek, Moonshot AI, Alibaba, and three other firms with illegally distilling model variants from Anthropic, OpenAI, Google, and SpaceX.

• Insilico’s AI-Designed Drug Shows Anti-Aging Signal: Trial data for rentosertib, an AI-designed drug for idiopathic pulmonary fibrosis, shows patients registered 2.7 to 3.5 years younger across six independent biological blood aging clocks.

• Anthropic Researcher Resigns Over Existential Risk: Pretraining researcher Jacob Coxon quits Anthropic, warning AI could trigger human extinction within a decade. Lead researcher Evan Hubinger backs Coxon, estimating a >10% probability of human extinction.

• DeepMind & Janelia Map Adult Fruit Fly Connectome: Researchers publish a complete wiring map of a fruit fly brain detailing 166,000 neurons and 125 million synapses reconstructed from electron microscope imagery using AI.

• Meta Launches “Muse” Personal Cloud Agent: Meta releases Muse, a personal agent that executes web tasks inside isolated cloud virtual machines using Stripe’s Link for single-use financial transactions.

• OpenAI Releases ChatGPT Images 2.5: OpenAI launches Images 2.5 featuring Sunburst and Flare models, reducing generation latency by 50% while adding stylus-to-image @Sketch capabilities.

• DeepMind Releases AlphaGenome Atlas: DeepMind publishes a free database predicting the functional impact of all 9 billion possible single-letter mutations across the human genome.

• OpenAI Research Agents Log 3.1 Workdays Per Human Day: Internal metrics show OpenAI researchers deploy 4+ agents simultaneously, with 90th percentile researchers spending over $7,000 daily in inference tokens.

• NBC Poll Shows 70% of US Voters Fear AI: A survey of 7,105 adults finds 70% feel more worried than excited about AI, with 69% opposing local data center construction.

OpenAI claims to have solved one of math’s “Millennium Problems” LINK

• OpenAI said its internal AI model proved the Navier-Stokes equations can “blow up,” a $1 million Millennium Prize Problem, but faces accusations it copied the approach from two outside mathematicians who got there first.

• The company said the proof took 88 hours and about 10,000 coordinating AI agents working in parallel, checked with Lean, software that verifies each step, producing a 100-page proof nobody outside OpenAI has reviewed.

• NYU’s Tristan Buckmaster says he told an OpenAI mathematician about his year-long project with Anthropic’s Levent Alpöge; days later, OpenAI’s Sébastien Bubeck said an internal model had a matching proof, which OpenAI denies stealing.

Meta launches a personal AI agent LINK

• Meta has launched Muse, a personal AI agent people can message to handle digital chores like sending emails, booking travel, or selling a car, all running inside a secure cloud setup the company says protects privacy from the start.

• Muse is out today on iOS and Android through a dedicated app, the Muse.ai website, and WhatsApp, with AI glasses support coming; it’s free to try, but heavy use needs one of Meta’s paid AI plans.

• The agent can buy things using Stripe’s Link tool, which issues single-use card numbers so real financial details stay private, and it isolates each user’s web activity in a virtual machine watched by a system Meta calls Sentinel.

US accuses Chinese firms of copying AI models LINK

• The U.S. government has accused six Chinese AI companies, including DeepSeek, Moonshot AI and Alibaba, of maliciously copying technology from American firms to train their own products more quickly and cheaply.

• Officials said the Chinese firms used a method called distillation, which trains smaller models using output from larger ones, to tap variants of models from Anthropic, OpenAI, Google and SpaceX, “likely with Chinese government awareness.”

• The accusations landed just before Chinese President Xi Jinping’s planned visit to the U.S. in late September, with officials warning the copying could boost China’s military and cyberattack strength against the U.S. and its allies.

Anthropic researcher quits over AI fears LINK

• Jacob Coxon, an AI researcher who worked at both OpenAI and Anthropic, quit Anthropic over ethics, warning that AI could kill countless people, possibly everyone, within the next decade if development keeps racing ahead.

• Coxon spent three years on pretraining research and accused both companies of “racing straight to self-improving superintelligence,” saying executives privately share the fear that AI could wipe out humanity even as they build it.

• Anthropic team lead Evan Hubinger backed Coxon, putting the chance AI kills all humans within ten years above 10%, and admitting the company has no clear plan yet to keep a superintelligence aligned with human safety.

Google mapped a fruit fly’s brain LINK

• Google and the Howard Hughes Medical Institute’s Janelia Research Campus published a complete wiring map of an adult male fruit fly’s brain on September 3, capping a decade-long project to trace every neural connection.

• The map, called a connectome, charts more than 166,000 neurons and 125 million synaptic links, making it the largest brain map by neuron count and a resource for studying how nervous systems perceive and react.

• To build it, scientists sliced the fly’s nervous system into millions of thin sections, imaged them with an electron microscope, and used AI to turn the 2D pictures into verified 3D reconstructions of the neurons.

ChatGPT image generation gets faster and sharper LINK

• OpenAI has launched ChatGPT Images 2.5, an update to its image-generation tool that produces pictures up to 50% faster than before while following instructions more closely across several rounds of edits.

• The update keeps details stable during edits, holding onto facial features, subjects and backgrounds the user didn’t ask to change, and it adds Sketch, a feature that turns finger or stylus drawings into reference points for generated images.

• It comes in two versions: GPT-Image-2.5 Flare, the default that makes images two to four times faster with better transparent backgrounds, and Sunburst, which is slower but sharper for product photos and marketing work.

OpenAI claims a Millennium Prize problem

Image source: OpenAI

The Rundown: OpenAI published a proof from an unreleased internal model, claiming to settle Navier-Stokes, one of the seven $1M Millennium Prize problems — with NYU’s Tristan Buckmaster questioning whether OAI raced with help from his drafts in Codex.

The details:

• OAI said it ran roughly 10,000 agents at once on a model “significantly more capable” than GPT-6 Astra, creating the proof in 88 hours.

• The company estimated the compute cost at “millions of dollars,” with Sam Altman calling it “one of the most amazing moments for me in OpenAI history.”

• Anthropic’s Levent Alpöge and Buckmaster spent a year on a similar route, feeding drafts into Codex and posting partial results the night before OAI’s own.

• Buckmaster released a statement saying OAI only started after hearing of their work, and never answered whether his Codex drafts trained the model.

• OAI said it “did not see any of their work” and that “no specific user data was accessed,” but can’t rule out that usage data may have improved its models.

Why it matters: The fight over credit has unfortunately overshadowed the biggest math breakthrough in an AI summer that has already changed the field. With an internal model already “significantly more capable” than the just publicly-released Astra, whatever ceiling you had in mind for what models can do this year probably needs to be raised.

Meta debuts new personal AI agent

Image source: Meta

The Rundown: Meta just introduced Muse, a new always-on personal AI agent with a text message-style interface that can handle tasks for users via its own cloud computer, like booking tables, sending emails, and shopping.

The details:

• Muse ties into apps like Gmail, Spotify, Ticketmaster, and OpenTable, with the agent also able to code its own integrations for ones that lack them.

• The agent works either in its own app or via WhatsApp, with a virtual machine that allows it to navigate websites, fill out forms, and take browser actions.

• Meta pushed privacy as a selling point, with Muse having its own secure cloud environment, Stripe-enabled payments, and built-in approval flows.

• Muse will have limited free usage and then monthly subscription tiers of $20 or $100, with the initial rollout only for U.S. users.

Why it matters: Between Hermes, OpenClaw, Grok Bot, Muse, and others, there is no shortage of personal agents hitting the market. But with frontier AI getting extremely good at navigating the web in their own browser and spinning up agents, part of us wonders if OpenAI and Anthropic aren’t far off from cutting out the agentic middlemen

OpenAI improves its already top-ranked image model

Image source: OpenAI

The Rundown: OpenAI just released ChatGPT Images 2.5, a new image model that cuts generation time by up to 50% over Images 2.0, brings upgrades to editing, and introduces features like sketch, templates, and shared images.

The details:

• OAI said 2.5 is “better at editing only what you’ve asked for,” avoiding the often awkward broader changes to the full image that have plagued previous models.

• Two models called Sunburst and Flare hit the API, which rank first and second on Arena AI’s Image leaderboards.

• Typing @Sketch lets users doodle-to-image, alongside new tools like comments for granular editing, templates for quicker starts, and shareable prompts.

Why it matters: A major math breakthrough in the morning, a new top-ranked image model by the afternoon. OAI’s shipping pace is heating up, and in some domains it’s starting to lap the field — Images 2.0 was still sitting at or near the top of the image leaderboards right up until today, so the only model OAI had to beat was its own.

ChatGPT may copy your writing style LINK

• OpenAI is testing a ChatGPT feature that studies your sent messages from connected apps so it can mimic your writing style and copy your tone in future responses.

• The tool pulls examples from linked services like Slack, Notion, Google Drive, and Gmail, sorted into Documents, Email, and Messaging, and may reach other apps through ChatGPT’s third-party tools.

• Only a select group of users has access right now, found under Settings, Personalization, Writing, with a toggle called “Reference my writing style”; OpenAI has not officially announced it yet.

DeepMind maps all 9 billion DNA mutations LINK

• Google DeepMind has used AI to predict the effects of all 9 billion possible single-letter changes to human DNA, releasing the results as a free database called AlphaGenome Atlas for academic researchers around the world.

• DeepMind built Atlas by running its AlphaGenome model across a reference human genome, generating about 27,000 predictions per variant covering gene activity and protein production across hundreds of cell and tissue types from humans and mice.

• The database includes a summary AVI score ranking each mutation’s impact, and early testers reported real gains, with one Exeter researcher finding 22% more genetic associations and reclassifying a variant tied to a patient’s epilepsy.

Tech firms plan AI data centers in Patagonia LINK

• Tech companies are eyeing Argentina’s Patagonia region for AI data centers worth billions of dollars, drawn by its cold weather, plentiful energy, and open land, though weak power and network infrastructure could stall the plans.

• Proposals include a $1.4 billion FlexDomes site and a $3 billion Green Capital project, while OpenAI signed a letter of intent for Stargate Argentina with Sur Energy, a renewable-powered facility that could reach 500 MW and $25 billion.

• The projects remain unconfirmed: OpenAI’s contract with Sur Energy is still unsigned, and Pampa Energia said interested parties want to start with small 20 MW to 40 MW pilots before scaling toward the full 500 MW.

TSMC bets on ASML machines to ease chip shortage LINK

• TSMC said this week it will start using ASML’s High-NA EUV lithography, a chip-printing method that relies on $400 million scanners, for high-volume manufacturing in 2030 after years of putting off the switch.

• The first High-NA tools will use standard 6×6-inch photomasks, which print finer 8nm features but cover only half the exposure area of today’s Low-NA systems, forcing chipmakers to stitch fields or split designs into chiplets.

• To fix that limit, TSMC is working with ASML on larger 6×12-inch photomasks, planning a pilot line in 2031 and production by 2033, an industry-wide change also backed by Intel and Samsung.

Google blames EU rules for worse search LINK

• Google reshaped its search results in Europe on Tuesday to obey a European Commission order, then told the public the new layout is worse, calling it the biggest quality drop in the engine’s 29-year history.

• The redesign puts one specialised search engine at the top, two more below it, and a carousel of hotels, airlines and restaurants with stripped-out features including real-time prices, a change Google says helps sites like Expedia and Booking.com over small local businesses.

• Google is the only source for its three key claims, including that earlier rules cut direct booking traffic by 30%, and it made the switch to avoid fines of up to 5% of worldwide turnover after a €460m penalty in July.

OpenAI opens the books on its AI research intern

Image source: OpenAI

The Rundown: OpenAI released a look at how much of its own research its coding agents handle, with agents logging 3.1 workdays for every one a human puts in, and the company hitting its “automated research intern” goal Sam Altman had set for September.

The details:

• Altman posted the timeline last October, calling for an intern-level system by this month and a fully automated researcher by March 2028.

• The typical OAI researcher now burns $600+ a day of agent tokens, with the 90th percentile above $7,000, and token output up 124x since December.

• Around 80% of researchers now use 4+ agents at a time, with OAI saying experiments per researcher are “at an all-time high” within the company.

• The agents’ tasks are also evolving, with more complex, longer tasks seeing greater success and areas like tech troubleshooting surging across the team.

Why it matters: OAI’s Tibo Sottiaux called internal use of Astra before its public release the “biggest competitive advantage”, and this internal usage report shows what its competitors are truly up against. The labs with unreleased frontier models and the budget to burn tokens on them are compounding in pretty unfathomable ways.

Insilico’s AI-designed drug hints at slowing aging

Image source: Insilico Medicine

The Rundown: Insilico Medicine published trial data showing that rentosertib, a drug its AI designed for a lung disease, left patients reading as biologically younger on all six “aging clocks” — an early signal (on a small sample) that it may help slow aging itself.

The details:

• Rentosertib treats idiopathic pulmonary fibrosis, where scarring slowly chokes the lungs, and Insilico’s AI picked the target protein and drew the molecule.

• The drug’s trial last year included 42 patients, with blood samples then used for the aging study and compared against a broader UK Biobank database.

• Six aging clocks (AI models by separate teams that read age markers from blood) rated treated patients as younger, with one group’s estimates falling 2.7–3.5 years.

• The dosing with the clearest aging signal was different than the one with the biggest lung-function gains, potentially showing effects beyond just disease relief.

Why it matters: Anthropic CEO Dario Amodei said that a real medical breakthrough can turn the tide on AI sentiment more than any marketing campaign could, and stories like this show the public what medicine may be about to become. We’re only at the starting line of a race of AI-designed drugs trying to reach the market.

U.S. voters’ rare bipartisan issue is fearing AI

Image source: NBC News

The Rundown: NBC News released a new poll of 7,105 adults on how Americans feel about AI, finding that 70% feel more worried than excited, and the concern spans both major parties — even as a growing share of Americans report using the tech.

The details:

• 52% use AI very often or sometimes, up six points from June 2025, but just 18% trust AI-generated information most or almost all of the time.

• Local data centers are a universal issue, with 69% against having one nearby and 45% strongly against, next to just 31% who would somewhat or strongly support.

• 70% believe AI is costing people jobs, and majorities expect it to hurt schools and elections, with science and healthcare as the only net positive wins.

• Washington’s AI rules fall short for 81%, and when asked which party they trust on AI policy, 44% say neither, with Democrats at 20% and Republicans at 16%.

Why it matters: AI sentiment has been on the mind of the industry’s top CEOs, but the polling is still moving in the wrong direction. Data centers have become the face of the issue, and the backlash is turning into a live campaign topic heading into the midterms. The other problem is that voters also don’t trust either party to handle it.

What Else Happened in AI on September 09th 2026?

OpenAI said an internal model more capable than GPT-6 Astra used 10,000 agents over 88 hours to produce a proposed solution to the Navier-Stokes Millennium Prize Problem. — CNBC

Suno replaced its models with Suno v6, trained on licensed music from Warner Music Group, BMG, and Believe. — TechCrunch

Meta launched Muse, a personal AI agent that runs in its own cloud VM with a built-in browser. — Axios

Apple’s “Surprise and Shine” event kicks off at 1 PM ET, where the company is expected to headline its first foldable, the iPhone Duo. — MacRumors

Google DeepMind launched AlphaGenome Atlas, a free, searchable map generated by its AlphaGenome AI that predicts what all 9B possible one-letter DNA mutations do.

Devin maker Cognition announced a new $2B funding round at a $48B valuation, with the company’s annualized revenue nearly doubling to $900M since May.

Inception Labs released Mercury 2.5, a diffusion model that writes text in parallel rather than word-by-word, claiming Haiku 4.5-level quality at 1,100+ tokens per second.

Mistral AI raised €3B ($3.5B) in new funding, valuing the French lab at over $24B, with the company pitching its models to organizations that want to control their own data.

U.N. rights chief Volker Türk said he shares fears that “advanced AI could pose an existential risk to humanity,” adding that a few men hold “almost unlimited power over AI.”

TikTok parent ByteDance is reportedly building a world model on top of its Seedance video AI, with founder Zhang Yiming running the project and a launch possible in Oct.

Chinese lab OpenBMB open-sourced MiniCPM5-2B, a new 2B-parameter model that ranks first among open models under 4B parameters on AA’s Intelligence Index.

OpenAI and Microsoft are facing a new lawsuit from the Seattle Times and Newsday over illegal training, with the outlets saying AI is breaking journalism’s economic cycle “beyond repair.”

🔗 RESOURCES

• DJAMGAMIND KIDS Bedtime Adventures: https://djamgamindkids.com

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow.