TL;DR
6 takeawaysAnthropic puts Claude Sheets in beta and makes Claude Data generally available on paid plans. [Anthropic]
OpenAI opens Businesses in ChatGPT in beta, with U.S. Business Agents for eligible ecommerce and online services. [OpenAI]
IBM and Microsoft announce Imagine Enterprise, an industry-specific agent platform delivered with consulting support. [Microsoft]
Short only: Google makes Gemini 4 Carbon available to paying developers, enterprises and AI Ultra subscribers. [Google]
Short only: DeepMind says AlphaProof Code formally verified key properties of Signal's protocol implementation. [Google DeepMind]
Short only: Claude Whiteboards enters beta as a shared board for real-time work with Claude. [Anthropic]
The Details
Story 01 / 03
Anthropic launches Claude Sheets in beta and takes Claude Data to general availability
- What happened
- On November 10 Anthropic announced Claude Whiteboards and Claude Sheets in beta, and said Claude Dashboards is now generally available under the name Claude Data. Sheets can generate spreadsheets with tabs, formulas, formatting and drop-downs, including cell-level edits with a change outline, and can be shared so a team works with Claude in the same file. Anthropic says sheet features are still simple for now, with export to .xlsx or CSV and a path into Claude for Excel or partner tools such as Airtable, Asana, monday.com and Smartsheet. Claude Data adds certified metric definitions (certified queries and live sharing remain in beta), data stories or narrated videos for review before sharing, live sharing that shows approved query results to colleagues without granting them connector access, design-system support, dashboard caching, and tracked changes. All three artifacts run on web, desktop and mobile; Data still needs a warehouse or CRM connector. They are available on all paid plans. On Team plans all three are on by default. On Enterprise plans, Whiteboards and Sheets are off by default, and Data turns on by default on November 19, 2026.
- Why it matters
- Anthropic is trying to keep boards, sheets and data work inside Claude instead of forcing a copy-paste handoff to another app. The announcement describes new product surfaces and admin defaults, not measured productivity gains.
- Enterprise impact
- Review Organization settings under Artifacts before a broad rollout. Certified queries and live sharing are still in beta. Live sharing shows only approved query results; viewers cannot run other queries with the sharer's access, according to Anthropic.
- Startup impact
- Teams without a warehouse connector can still try Sheets. When Excel-level features are required, Anthropic points to export or Claude for Excel rather than treating the beta sheet as a full spreadsheet replacement.
- What to focus on
- Keep Whiteboards as a short item only; this card is about Sheets and Data. Confirm plan defaults and the November 19 Enterprise Data default before assuming every workspace has the features on.
Discuss this story
Comment with your GitHub account. One thread per story.
Story 02 / 03
OpenAI introduces Businesses in ChatGPT in beta
- What happened
- On November 10 OpenAI launched Businesses in ChatGPT in beta. Through Business Hub, eligible businesses in markets where ChatGPT is available can claim and verify their organization, update details, and add self-service product feeds with pricing and availability. Shopify merchants can connect a store from Business Hub or the Shopify admin; OpenAI says eligible Shopify Catalog products are already integrated. Business Hub also offers aggregated insights such as mention counts, website clicks and discovery topics. OpenAI says it does not share user conversations or Business Agent chats with businesses, and that insights appear only when volumes meet privacy thresholds. After a claim, eligible ecommerce and online service businesses in the United States can build, test and launch a Business Agent. When a business appears in a model response, a person can open its ChatGPT profile and, if an agent exists, start a clearly labeled conversation that OpenAI says is separate from the original chat and from ChatGPT's independent answers. Businesses can also sponsor a Business Agent as a destination for ChatGPT Ads; sponsored placements are described as clearly labeled and separate from ChatGPT responses. OpenAI cites more than 1.2 billion weekly ChatGPT users as the audience figure.
- Why it matters
- This is a discovery and customer-conversation surface inside ChatGPT, not a general-purpose agent API. Eligibility, labeling and the separation from ChatGPT's own answers are the operational details to verify.
- Enterprise impact
- Existing OpenAI Enterprise customers and advertisers can claim through their representative. Treat the weekly-user figure and privacy-threshold design as vendor statements. Confirm whether your category and country qualify for Business Agents; the agent feature is limited to eligible U.S. ecommerce and online service businesses at launch.
- Startup impact
- Shopify merchants get a shorter verification path. Sponsored agent placements are an advertising product, not the default organic profile.
- What to focus on
- Do not imply global Business Agent availability. OpenAI says it plans to expand to more markets and categories over time; that is a stated intention, not a shipping date.
Discuss this story
Comment with your GitHub account. One thread per story.
Story 03 / 03
IBM and Microsoft announce Imagine Enterprise for industry-specific agents
- What happened
- On November 10 IBM and Microsoft announced Imagine Enterprise, a platform meant to help enterprises build, run and govern AI agents for industry-specific processes, with particular emphasis on regulated industries. IBM Consulting contributes what it describes as more than 6,000 prebuilt AI agents and assets on IBM Consulting Advantage, plus industry expertise. Microsoft supplies the AI, data, cloud and security foundation, including technologies the announcement names such as Microsoft Foundry, GitHub Copilot, Microsoft Copilot Studio, Fabric, Entra, Purview, Defender and Azure. Forward-deployed engineers from both companies are offered for qualifying engagements. Initial industry frameworks cover financial services, insurance and retail, with other industries described as possible later. The platform is available by engaging IBM Consulting to deploy and operate it, by co-innovating on enterprise-scale implementations, or through self-management options. SEI is cited as exploring agentic technologies with IBM Consulting and Microsoft; that is a customer illustration, not a published production result. The release notes that statements about future direction are goals and may change.
- Why it matters
- The pitch is a consulting-delivered agent layer on existing enterprise systems, not a self-serve model drop-in. The 6,000-asset figure and governance claims come from the companies.
- Enterprise impact
- Ask what is actually included in a given engagement, how agents connect to current systems, and which controls are contractual versus planned. Self-management is mentioned, but the primary path described is IBM Consulting deployment or co-innovation.
- Startup impact
- This announcement is aimed at large, often regulated, buyers. It is not a public API launch for independent builders.
- What to focus on
- Attribute the prebuilt-agent count to IBM. Do not treat the SEI mention as evidence of completed production outcomes.
Discuss this story
Comment with your GitHub account. One thread per story.
Worth a Glance
Gemini 4 Carbon is generally available: midday recap
Google's November 10 post makes Gemini 4 Carbon available via the paid Gemini API in AI Studio and Gemini for business, and is rolling it out to Google AI Ultra subscribers in Antigravity, the Gemini app and AI Mode in Search. Google calls Carbon its strongest model for long-horizon agent work and cybersecurity, and says Argon remains the better default for most software engineering and knowledge work because Carbon uses more tokens. Vendor-reported ranks include #1 on CursorBench 4.0, Perplexity's WANDR and Zapier's AutomationBench at more than 2x lower cost per task than the next best external model. Introductory pricing matches Argon at $2 per million input tokens and $10 per million output tokens through November 30, 2026, then $4 and $20. Antigravity is previewing /swarm, which Google describes as turning one prompt into a coordinated team of agents. Midday post only.
Google(opens in a new tab)DeepMind AlphaProof Code: morning recap
Google DeepMind said researchers used AlphaProof Code, with Gemini Carbon agents and human-reviewed specifications, to prove key security properties of the actual Signal protocol implementation. DeepMind reports over six million lines of Lean, human review of about one percent of the codebase, a four-week effort it contrasts with years of traditional work, and machine-checked proofs that two reintroduced, previously fixed bugs were insecure. Novelty and historical-time claims are the lab's. Morning post only.
Google DeepMind(opens in a new tab)Claude Whiteboards enters beta: short recap
Anthropic's same November 10 article puts Claude Whiteboards in beta on paid plans. It is a shared board for real-time brainstorming with Claude, including turning a messy plan into a timeline with owners and dates, grouping retro notes, and exporting to a doc, deck or partner tools such as Lucid, Miro or Vani by Zoho. Whiteboards was already posted separately today; this is a brief recap only.
Anthropic(opens in a new tab)
One Thing Not to Miss Today
Keep today's model launch in the short column
Gemini 4 Carbon is the largest model drop of the day, but it was already posted at midday, so it stays a brief recap here. The new details worth a closer look are where work artifacts now live inside Claude, and how businesses will claim space inside ChatGPT.