01Lead
Learn how OpenAI scaled Habitat with reliable multi-tenancy, global caching, and Azure Cosmos DB to deliver 12× faster reads for ChatGPT.
OpenAIModels
— — —
15/15 wires up
Live
OpenAIRapidly scaling online storage to serve over 1 billion ChatGPT users (Part II)Hugging FaceLitPivot: Developing research ideas through dynamic contextualization within the literature landscapeThe VergeMeta delays new business terms after publisher pushback on AI scrapingThe VergeRetro gaming is being overrun with AITechCrunchUber rival inDrive eyes 2027 robotaxi tests in Central Asia with Pony AI and NvidiaNVIDIAKazakhstan Expands Compute for Its AI FutureTechCrunchLeft alone with nothing to do, AI agents started playingOpenAIOpenAI and Poland’s NASK partner to advance AI safety and securityHugging FaceDoes Your Model Run on AWS Trainium? Look It UpMIT Tech ReviewDesigning an AI serving harness built to lastWIREDThe Social Society Helping Women Join the AI BoomMIT Tech ReviewThe Foundations of Scalable AI
Free daily digest
The AI news that matters, with the 'why it matters' for builders, founders and execs.
5–10 hand-picked items a day from the labs and the wires. Every item links to its source and says why it matters to your work. Five minutes, no hype.
Daily digest
Friday, October 23, 2026
OpenAI eyes a $1.25T valuation, Anthropic and DeepSeek make agent and funding moves, and AI privacy terms get rewritten
- 01
OpenAI working on new $30B raise at $1.25T valuation ahead of IPO
Source: TechCrunch
TechCrunch reports OpenAI is working on a new $30 billion raise at a valuation above $1 trillion, which would make it the most highly valued private company ever.
Why it mattersIf you build on OpenAI, your vendor is turning into a public-market company, and that usually means pricing, packaging and roadmap get judged on margins. It is worth pressure-testing how locked in you are, and keeping a second model provider warm.
- 02
Build more responsive apps with the Decisions API
Source: OpenAI
OpenAI announced a Decisions API, pitched as fast, focused, low-cost decisions for any application. The announcement is short; TechCrunch frames it as a competitor in a growing field of decision models.
Why it mattersA lot of product logic is routing, triage and yes/no calls that do not need a frontier model. If this is as cheap and fast as the pitch suggests, it is a candidate for the boring 80% of your AI spend. Check the docs and benchmark on your own data before you rewire anything.
- 03
Claude Duet app is Anthropic’s bet on agents
Source: TechCrunch
Anthropic’s Claude Duet app is its push into agents, and per TechCrunch it is available through a waitlist for Pro and Max plan users.
Why it mattersThe agent race is moving from APIs into products people use directly, which squeezes startups that sell a thin agent layer over someone else’s model. If you build agent software, be clear about what you do that a lab’s own app will not.
- 04
DeepSeek reportedly raises $15B days after US restrictions
Source: TechCrunch
TechCrunch reports, citing unconfirmed reporting, that DeepSeek raised $15 billion in one of China’s largest funding rounds, days after the U.S. ordered federal agencies to remove its technology over national security concerns.
Why it mattersUS restrictions have not cut DeepSeek off from capital, so expect it to remain a serious low-cost model competitor. If you sell to US government or regulated customers, check whether any DeepSeek-based components sit in your stack.
- 05
1Password launches a way for AI agents to log in on your behalf without seeing your password
Source: TechCrunch
A new 1Password feature lets agents sign in to a customer's personal accounts while keeping the underlying credentials hidden from the agent.
Why it mattersHanding agents your passwords is the quiet blocker for letting them do real work. A pattern where the agent never sees the secret is the shape security teams will ask for, so it is worth designing for now.
- 06
Donating Agent Skills to the Agentic AI Foundation
Source: Simon Willison (quoting Anthropic)
Anthropic is donating the Agent Skills standard to the Agentic AI Foundation under the Linux Foundation, the same home as the Model Context Protocol. Simon Willison notes he had predicted this move.
Why it mattersNeutral governance lowers the risk of building on a shared format for packaging agent abilities. If you are investing in skills or MCP tooling, this is a signal that those formats are meant to outlive any one vendor.
- 07
Microsoft’s Mustafa Suleyman says its AI outperforms doctors on hard cases as he chases ‘medical superintelligence’
Source: TechCrunch
Microsoft’s new AI diagnostic system beat generalist doctors on hard cases in a study published in Science, but TechCrunch notes most of the improvement appears to come from the underlying AI models rather than Microsoft’s system itself.
Why it mattersHealthcare is where AI claims face the toughest scrutiny, and the detail about where the gains come from is the one to watch. For anyone building or buying clinical AI, ask what the wrapper adds beyond a better base model before paying for it.
- 08
Meta turns Muse sharing back on
Source: The Verge
Meta re-enabled sharing in Muse after researchers reported that public conversation links were easy to find, including pages like a company's internal revenue review and a small business owner's loan application.
Why it mattersAny 'share this AI conversation' feature is a data-leak surface. If your team pastes sensitive material into AI tools, assume shared links can be discovered, and if you ship such a feature, make links unguessable and expiring by default.
- 09
Now Google says it might use your Gmail and Google Photos data to target ads
Source: The Verge
Google updated its privacy policy to add an exception to its promise not to use Drive, Gmail or Google Photos for personalized ads, which could apply if you use personalized AI features. The change was spotted by Android Authority.
Why it mattersPersonalized AI features are paid for with data permissions. Execs rolling out AI inside Google Workspace should re-read what is switched on for their teams, and anyone building on user data should expect this question from customers.
- 10
Meta delays new business terms after publisher pushback on AI scraping
Source: The Verge
Meta is delaying enforcement of its new business terms, now effective November 20th, after publishers worried the terms could let it scrape news sites for AI training.
Why it mattersWho may train on what is being fought out in contract language, not just courtrooms. If you publish content or rely on Meta's platforms for distribution, read the terms before the new date; if you train models, expect more scrutiny of your data sources.
Live desk
Everything moving on the wires right now, ranked automatically. The daily digest above is the human-curated version.
03
Should OpenAI command an over $1 trillion valuation with this round, it would break records as the most highly valued private company ever.
04
The new feature allows agents to log into a customer's personal accounts while still protecting the private credentials that secure those accounts.
05
Jack Dorsey's Block announced an AI model called Dessa that's trained "from scratch on nothing but human conversation," and to show off the model, Block shared chats it has had with humans that can occasionally be a little awkward and uncanny. Take this one, for example, shared by Dessa lead Ragavan Thurairatnam. The…
Index
06
Hosted by TechCrunch in partnership with Fidelity Private Shares, the full-day event goes beyond fundraising and AI to explore the realities of finding customers, building teams, sharpening a pitch, developing founder leadership, and scaling companies both inside and outside traditional startup hubs.
08
Google has updated its privacy policy in a way that could mean you'll start seeing targeted ads based on your personal data if you're using personalized AI features. The update, spotted by Android Authority, adds an exception to its promise about not using your Drive, Gmail, or Google Photos for personalized ads.…
09
Meta is delaying the enforcement of its new business terms after publishers raised concerns that they could allow the company to scrape news sites for AI training, as reported earlier by Fast Company. In a statement to The Verge, Meta spokesperson Daniel Roberts says the terms will go into effect on November 20th "to…
10
OpenAI explains how we strengthened containment for frontier model training and evaluation, including offline environments and high-compute red-teaming.
OpenAIModels11
Launched today, NVIDIA Minerva is an AIOps platform built to help teams find problems earlier and keep AI factories productive. It brings together telemetry from infrastructure and AI workloads, then applies purpose-built AI models to turn that data into timely, actionable insight. AI factories are among the most…
12
Retro video games are everywhere lately - especially in places where they aren't supposed to be. Halo in a web browser, Wind Waker on the iPhone, Super Smash Bros. in mixed reality. Over the past month or so, these kinds of unexpected projects have gone viral with incredible frequency thanks to new AI tools, as […]
14
OpenAI and Poland’s NASK are partnering on AI safety, security, and technical research to strengthen safeguards and support a more resilient digital ecosystem.
OpenAILabs15
The next competitive advantage in enterprise AI has less to do with which model you choose than with what you build around it. That’s an important shift. For the past several years, much of the AI conversation has centered on model capability: bigger models, better reasoning, larger context windows, higher benchmark…
17
A global map categorizing every 10-meter patch of land on Earth into distinct forest types.
DeepMindLabs18
A scalable AI system requires a strong foundation. Data, engineering, platforms, operating models, and governance must converge to work as one. This session from MIT Technology Review’s EmTech Future conference explores what it takes to build that foundation in a highly regulated environment and how a responsible,…
21
WIRED spoke with the founders of the Aged Intelligence Society, a fast-growing events startup that brings women together to “solve midlife problems using frontier tools.”
22
Enterprise AI can generate recommendations, automate tasks, and help organizations make decisions faster. But without an understanding of how a business actually operates, even powerful models can miss the constraints and relationships that determine whether a decision is the right one. As companies look to move AI…
23
Ask most executives to picture the cost of AI and they’ll likely see GPUs—racks of accelerators with gigawatts of power—and a waitlist for silicon. But that picture leaves something out: Data storage costs. What happens to data after a model responds? Much of the back-end data that’s generated—from billions of daily…
24
Kazakhstan’s AI ecosystem is moving from infrastructure plans to deployment. The shift is visible across the full AI stack — from infrastructure to models to applications. The announcements below — coinciding with a visit to Astana by NVIDIA founder and CEO Jensen Huang — point to the nation’s increasing supply of…
25
Step one in convincing people to live with robots with cameras on their faces? Disguise them as cute home decor.
26
<p><strong><a href="https://www.kimi.ai/blog/kimi-k3-1-pro">Introducing Kimi K3.1 Pro</a></strong></p> Kimi K3.1 Pro landed on the Kimi API today - it's not yet open weights, but Moonshot promise "Full model weights will be released by Nov 4, 2026".</p> <p>They're also…
Simon WillisonModels27
The new open-source model speaks the language of DNA and can help researchers understand the effects of any changes on gene activity and health.
28
The tool, which is not yet widely available, will compete with OpenAI’s dots and Meta’s Muse. It is built for work and not personal errands.
29
Amazon Web Services logged attempts to leverage the internet-melting PHP2Shell bugs four days ago. They’re now being “broadly exploited” across the internet.
30
Hugging FaceResearch31
At GTC Berlin, banks, payment companies and financial data providers are showing what production-scale intelligence looks like — built on NVIDIA accelerated computing, powered by open weight models and running on their own infrastructure. Revolut is rebuilding its technology stack on a single foundation model PRAGMA,…
32
To help developers rapidly build world models, which help physical AI systems understand and simulate the real world, NVIDIA has expanded the Cosmos platform with new training tools, simulation checkpoints and Cosmos Data Factory blueprints. Ranking top five on six different open models leaderboards, the NVIDIA Cosmos…
33
<p><strong><a href="https://www.anthropic.com/news/donating-agent-skills-to-the-agentic-ai-foundation">Donating Agent Skills to the Agentic AI Foundation</a></strong></p> News today from Anthropic:</p> <blockquote> <p>We’re committed to keeping Agent Skills open,…
Simon WillisonAgents34
<blockquote cite="https://twitter.com/mitsuhiko/status/2113207589978636646"><p>As fun as AI is, we built so much bad software over the last 30 years, and now the universe is punishing us for our sins.</p></blockquote> <p class="cite">&mdash; <a…
Simon WillisonResearch35
Hugging FaceResearch36
Hugging FaceResearch37
Hugging FaceResearch