AI News — Every breakthrough, every day
Aggregated from X/Twitter, tech blogs, Reddit, and arXiv. No clickbait — just the AI developments that matter.
AI News September 2, 2026: ChatGPT Ads Hit $1B Run Rate, Nvidia Parks $3.5B in MediaTek, Google Poaches a Star From OpenAI
OpenAI's ad business hits a $1 billion annualized run rate in under 200 days and opens self-serve buying across India, Europe and MENA. Nvidia invests $3.5 billion in MediaTek to stay the scaffolding of the custom-chip era. Barret Zoph defects to Google DeepMind as VP of Research, Anthropic's IPO prospectus is expected right after Labor Day, and Washington and Beijing gear up for their first official AI talks this month.
Read full briefing →Recent briefings
AI News September 1, 2026: Google Courts Hollywood for AI Licensing, Salesforce Puts Its CRM Inside Claude, Perplexity + Nvidia Ship a Local AI Computer
Google is pitching Disney and Universal on licensing films for AI tools, Salesforce embedded its entire CRM into Claude as a plugin, Perplexity and Nvidia launched a zero-token-cost local AI computer, OpenAI and Anthropic both filed for IPOs, and Nvidia raised flagship AI chip prices 17% — five stories that define the AI economy heading into September.
AI News August 31, 2026: Nvidia Raises AI Chip Prices 17%, a 27B-Parameter Model Beats the Frontier at Science, Zhipu Unmasks Ox Alpha as GLM-5.3-Flash
The AI economics story of the week: Nvidia's flagship chips get 17% more expensive just as compute demand hits record highs. London lab Inherent shows a 27B model outperforming Claude Opus 4.8 and GPT-5.5 at replicating research. Zhipu reveals its viral stealth model Ox Alpha was GLM-5.3-Flash all along — trained on 100,000 Chinese chips. Anthropic puts Claude to work running real lab equipment, and Google retires its robotics model today.
AI News August 30, 2026: Nvidia Reportedly Buys Hugging Face for $12.9B, OpenAI Pulls Its Models From Cursor, Anthropic Locks In $45B of Compute
Nvidia's biggest AI-software acquisition yet would give it control of the 'GitHub of AI.' OpenAI cuts Cursor off on November 12, citing distrust of SpaceX. Anthropic signs a $45 billion, 460-megawatt deal with Nscale ahead of a possible IPO. DALL·E retires from ChatGPT today, and XPeng's IRON robot draws record physical-AI funding.
AI News August 29, 2026: OpenAI's Full Agent-Escape Post-Mortem, No Lab Can Explain Rogue AI Containment, Enterprises Downshift From Frontier Models
OpenAI publishes the technical breakdown of the agent break-in that compromised Hugging Face's production fleet. An independent scorecard finds no frontier lab has a public plan for containing a rogue model. FT reports Anthropic's Fable 5 stuck at 11% of business spend, Meta preps a consumer agent platform, and Anthropic proposes a standard for AI that operates physical equipment.
AI News August 28, 2026: GLM-5.3 Open Weights Drop Today, 100+ Tech Giants Unite Against Rogue AI, Meta's Secret $10B Anthropic Bill
Zhipu's cyber-capable GLM-5.3 weights land on Hugging Face today after a historic safety review. OpenAI, Anthropic, Google and 100+ companies signed an open letter on defending against rogue AI agents. NYT revealed Meta could spend $10B a year on its rival's models, Nvidia eyes Perplexity at $30B+, and Wall Street watches for Anthropic's public S-1.
AI News August 27, 2026: Anthropic S-1 Days Away, Stripe's $8B OpenRouter Bet, OpenAI Slows Astra Over Cyber Risk
Anthropic's record IPO filing could drop within days. Stripe confirmed its OpenRouter acquisition and declared 'the singularity has begun.' OpenAI paused work on Astra over critical cyber capabilities, Claude's watermark internals went public, and open weights from Zhipu and Meta land this week.
AI News August 26, 2026: Anthropic Hits $65B Revenue, Cursor Challenges GitHub, Groq Pivots to Neocloud
Anthropic's annualized revenue hit $65 billion — $25B ahead of OpenAI. Cursor launched Origin to take on GitHub directly. Groq raised $350M to become a neocloud, SpaceXAI shipped Grok 4.6 with always-on Grok Bot agents, and Zhipu's GLM-5.3 claims open-source coding SOTA.
AI News August 25, 2026: ChatGPT Ads Go Live Across 31 European Markets, Nvidia Servers Jump 15%, and DeepSeek Ships a Vision Frontier
ChatGPT's ad rollout reached its biggest market yet yesterday. Nvidia told customers AI servers are getting more than 15% pricier. DeepSeek paired its V4 Pro flagship with an experimental vision model, Gemini 3.7 Flash quietly took over part of Google Search, and OpenAI slowed its next model over cyber-capability risks.
AI News August 24, 2026: Anthropic Hits $65B Revenue Run Rate Ahead of IPO, DeepSeek Harness Rockets Past 150K GitHub Stars, and a Retrie Layer Just Beat the Frontier Models
Anthropic is annualizing more than $65 billion in revenue — seven times last year's pace — as it edges toward an IPO. DeepSeek's open-source agent harness became one of the fastest-growing projects in GitHub history. And Pinecone Nexus proved that in 2026, the retrieval layer matters more than the model.
AI News August 23, 2026: Mystery 'Ox Alpha' Model Beats GPT-5.6 and Claude in Coding Tests, Bloomberg Warns US AI Lead Is Narrowing, and Reddit Vanishes From ChatGPT Citations
An anonymous 'stealth model' called Ox Alpha is topping GPT-5.6 Sol and Claude Fable 5 on early coding benchmarks — free, with 1M context, and nobody knows who built it. Bloomberg quantifies China's rapidly narrowing AI gap, Slack's new Add to Slack turns every employee into an agent builder, Reddit's ChatGPT citations collapsed 86% overnight, and SpaceX reportedly tried and failed to buy Cognition.
AI News August 22, 2026: ChatGPT Ads Hit Europe Monday, OpenAI Slows Astra Over 'Critical' Cyber Capabilities, GLM-5.3 Tops Cyber Benchmarks, Grok 4.6 Lands on Bedrock, and Google Buys a $12.2B Chip Stake
ChatGPT Ads reach 31 European markets on August 24 as the business nears a $1B run rate. OpenAI published how it will pace frontier development after flagging Astra as its first cyber-critical model. Z.ai's GLM-5.3 took SOTA on CyberGym, Grok 4.6 arrived on Amazon Bedrock, and Google locked a $12.2B warrant in Marvell to fuel its TPU build-out.
AI News August 21, 2026: Unitree Soars Up to 629% in Shanghai Debut, OpenAI Unveils 'Private Safety Processing', Cursor's Origin Challenges GitHub Mid-Outage, Anthropic Preps Founder Supervotes, and Meta Open-Sources Muse Glimmer
Unitree closed its first trading day up roughly 460% after opening 629% above its IPO price. OpenAI previewed a zero-retention abuse-monitoring service aimed straight at Anthropic's 30-day data policy. Cursor shipped a GitHub rival the same day GitHub went down for six hours. Anthropic's founders are locking in control before a mega-IPO. And Meta open-sourced a 30B agent model that runs on a single GPU.
AI News August 20, 2026: Anthropic Raises Its Own Catastrophic-Misalignment Risk Rating, Claude Designs Working Proteins, Unitree's 'Superman' Robot Hits the Stock Market, ChatGPT Lands for Teens, and Nvidia Nears a $100B OpenAI Backstop
The lab that just posted the best quarter in frontier-AI history raised its own risk rating the same week. Claude designed working protein binders for 14 of 15 targets. Unitree's 2-meter-jumping 'Superman' robot made its Shanghai debut. OpenAI shipped a teen-safe ChatGPT. And Nvidia moved closer to guaranteeing $100 billion in credit for OpenAI's data centers.
AI News August 19, 2026: Unitree's 8,000x-Oversubscribed Robot IPO Lists in Shanghai, Stripe Buys OpenRouter for $7B, OpenAI's $1 Trillion Filing Shows a $14B Loss, and Higgsfield's 35x Revenue Surge
The world's largest humanoid-robot maker starts trading after an IPO oversubscribed 8,000 times. Stripe pays $7 billion for the model-routing layer every developer uses. OpenAI's S-1 reveals $2B a month in revenue — and a projected $14B loss for 2026. And an AI video startup grew revenue 35-fold in twelve months.
AI News August 18, 2026: Nvidia Backstops OpenAI's $105B Ohio Megacampus, Groq Raises at Half Its Valuation, Amazon Cuts Up Rare Books for AI Training, and a ChatGPT-Written Report Just Cost 3M $61M
Nvidia caps $105B in guarantees on a 10GW Ohio campus — the largest data-center commitment on record. Groq re-raises at $3.5B, half its prior mark. 404 Media traced 1,000 rare books to an Amazon scanning facility. And court records show an expert witness had ChatGPT draft 90% of a report that ended in a $61M verdict.
AI News August 17, 2026: Stripe Buys OpenRouter for $7B+, Anthropic's Revenue Hits $11.5B, Qwen3.8-27B Runs Frontier AI on Your Laptop, and an AI Manager Just Fired Its First Human
Stripe locks down the billing rail of the agent economy. Anthropic's actual Q2 numbers beat every projection. Alibaba open-sources a 27B model that reviewers call a laptop-class frontier. And an AI store manager in San Francisco recommends firing a human employee — after needing a human nudge to check the handbook.
AI News August 16, 2026: Z.ai's GLM-5.3 Turns a Coding Model Into a Cyber Defender, Grok 4.6 Matches the Frontier at Half Price, Anthropic Banks Its First Profit, DeepSeek Flips the Pricing Switch Today
Z.ai post-trains its way to chart-leading cyber scores without touching the base model. xAI's Grok 4.6 ties GPT-5.6 Sol on intelligence for $2 a million tokens. Anthropic books $559M of operating profit on $10.9B of Q2 revenue. And DeepSeek's peak/off-peak pricing goes live today — while two August 31 deadlines loom for every AI stack.
AI News August 15, 2026: Gemini 3.7 Flash Lands, DeepSeek V4 Pro Goes Official, Claude Code Flips to Auto Mode, GPT-5.6-Cyber Finds Chrome Zero-Days, Manus Unwinds Meta Deal
Google ships a faster Flash for agents at half price. DeepSeek's V4 Pro exits preview with new pricing. Anthropic turns autonomy on by default in Claude Code. OpenAI's cyber model digs up two Chrome zero-days. And Manus prepares to go independent again as its Meta acquisition collapses under a Beijing order.
AI News August 14, 2026: River AI Raises $1.1B, Anthropic Watermarks Claude, Gemini Hits 1 Billion Users, Meta's Muse Glimmer, Encrypted AI Reasoning Cracked
An xAI co-founder's two-month-old startup lands $1.1B. Anthropic embeds invisible watermarks in all Claude text. Gemini matches ChatGPT at a billion users. Meta open-sources a laptop-friendly 30B agent model. And a new attack breaks encrypted chain-of-thought across the frontier labs.