Businesses in ChatGPT went live in beta on 10 November, and it is the most important product OpenAI has shipped this week that is not a model name. In its official announcement, OpenAI said companies can now claim a profile in ChatGPT through Business Hub, keep their information current, upload product feeds with pricing and availability, and see aggregated insights on how people engage with them. Eligible U.S. ecommerce and online-service businesses can also launch a Business Agent — a labeled, separate conversation that helps customers explore an offering inside ChatGPT. OpenAI puts the audience at 1.2 billion people who use ChatGPT each week. That figure is the company’s claim, not a third-party audit.
The same 48 hours delivered a rival workbench. On 10 November, Anthropic began rolling out Claude Sheets and Claude Whiteboards in beta, made Claude Data generally available, and expanded interactive connectors so partner apps can open inside Claude. Today, 12 November, a new Anthropic usage policy also takes effect. This is a Cross Analysis, not a launch recap: the frontier labs are no longer only racing at the model layer. They are racing to become the place where work, and now commerce, actually happens.
What Businesses in ChatGPT gives a company — and what it withholds
The claim flow is straightforward. Any eligible business in a market where ChatGPT is available can claim its organization in Business Hub and, once verified, update key details and add self-service product feeds. Existing OpenAI Enterprise customers and advertisers can work through their representative. Shopify merchants can connect from Business Hub or the Shopify admin; OpenAI says eligible Shopify Catalog products are already integrated into ChatGPT.
Insights are deliberately thin, and that is the privacy story. Business Hub, OpenAI says, will show aggregated mentions, website clicks, and “discovery topics” — common themes in conversations where a company’s website is cited. OpenAI says it does not share users’ conversations, or Business Agent chats, with businesses. Insights appear only when volume clears a threshold. That is a tighter wall than a typical ads dashboard, and it is also a reminder that the company is trying to keep ChatGPT feeling like a trusted advisor, not a storefront with a two-way mirror.
The Business Agent is the sharper edge. When a business appears in a model response, a user can open its ChatGPT profile and, if an agent exists, start a conversation that OpenAI says is clearly labeled, kept separate from the original chat, and distinct from ChatGPT’s own answers. Businesses can also sponsor that agent as a destination for ChatGPT Ads, with sponsored placements labeled. For now, agents are limited to eligible U.S. ecommerce and online-service categories. OpenAI says it plans to expand to more markets and more kinds of business. No public price for the agent, or for sponsorship, was in the announcement. We will not invent one.
Claude’s workbench, and the policy that kicked in today
While OpenAI built a storefront inside the chat box, Anthropic spent 10 November turning Claude into a place where the work file already lives. Reporting on the launch, including The Neuron and TLDR AI, describes Claude Sheets and Claude Whiteboards rolling out in beta on paid plans, across web, desktop and mobile, and Claude Data — previously Claude Dashboards — moving to general availability, with connectors into warehouses and CRMs. Interactive connectors, in the same wave, open partner tools such as Adobe, Atlassian, Box and Notion inside Claude. Claude Meetings, described as a notes-and-documents presence in calls, is still ahead.
The trust layer moved on a different clock. Anthropic’s updated usage policy, announced on 8 October, takes effect today, 12 November. Coverage of the policy, including SH Consulting’s 12 November briefing, says the most visible addition is a ban on sustained, needless abusive or cruel behavior toward the models, with ordinary frustration, dark creative work and model testing carved out, and a tighter single section on deceptive campaigns and artificial activity. That is a governance story running in parallel with a product story. One lab is inviting businesses into the chat. The other is tightening what the chat may be used for, on the same week it dropped spreadsheets and whiteboards into the product.
Neither move replaces the polymath. A Business Agent still needs a human who can stand behind the answers. A Claude Sheet still needs someone who knows which number is allowed to ship. That is the through-line of What Makes an AI Polymath?, Why Global Corporates Are Competing for AI Visionaries, and The New Multi-Hyphenate Standard. The studios already working this way are the ones we described in The Multi-Hyphenate Advantage.
The Cross Analysis, if you are building or booking in this window
OpenAI is commercializing discovery: claim, feed, agent, optional ad. Anthropic is commercializing production: sheet, board, data, connector, with a stricter rulebook taking effect the same week. Google, in the same 48-hour window, put Gemini 4 Carbon and /swarm into the API. The labs are building three different operating systems around the same user. The exclusive move is not to pick a mascot. It is to keep a human who can work all three without becoming the campaign the new policies were written to stop.
If you are hiring that kind of public intelligence for a stage or a private briefing, start at Book a Celebrity and the Celebrity Fee Meter. The pace of this week is a vitality problem, which is why Everyday Vitality by Richiegaps and Motivations Masters’ Orientation, Nothing is 100% belong here. To enter the Cup, go to Register Interest. Businesses in ChatGPT is the storefront. Claude is the desk. The Cross Analysis is that both showed up in the same 48 hours, and neither one runs itself.